Prove Fix Skill | Agent Skills

Codex compatibility note:

Invoke repository skills with $skill-name in Codex; this mirrored copy rewrites legacy Claude /skill-name references.

Task tracker mandate: BEFORE executing any workflow or skill step, create/update task tracking for all steps and keep it synchronized as progress changes.

User-question prompts mean to ask the user directly in Codex.

Ignore Claude-specific mode-switch instructions when they appear.

Strict execution contract: when a user explicitly invokes a skill, execute that skill protocol as written.

Subagent authorization: when a skill is user-invoked or AI-detected and its protocol requires subagents, that skill activation authorizes use of the required spawn_agent subagent(s) for that task.

Do not skip, reorder, or merge protocol steps unless the user explicitly approves the deviation first.

For workflow skills, execute each listed child-skill step explicitly and report step-by-step evidence.

If a required step/tool cannot run in this environment, stop and ask the user before adapting.

Codex Project-Reference Loading (No Hooks)

Codex uses static project-reference loading instead of runtime-injected project docs. When coding, planning, debugging, testing, or reviewing, open project docs explicitly using this routing.

Always read:

docs/project-config.json (project-specific paths, commands, modules, and workflow/test settings)
docs/project-reference/docs-index-reference.md (routes to the full docs/project-reference/* catalog)
docs/project-reference/lessons.md (always-on guardrails and anti-patterns)

Missing/stale context route: If docs/project-config.json, the docs index, lessons.md, CLAUDE.md, AGENTS.md, or any task-required reference doc is missing or stale, auto-run $project-init or the narrow setup route ($project-config, $docs-init, $scan-all, $scan --target=<key>, $claude-md-init) before ordinary project-specific work. If Codex mirrors or AGENTS.md are missing/stale, ask the user to run $sync-codex; do not auto-run it.

Situation-based docs:

Backend/CQRS/API/domain/entity changes: backend-patterns-reference.md, domain-entities-reference.md, project-structure-reference.md
Frontend/UI/styling/design-system: frontend-patterns-reference.md, scss-styling-guide.md, design-system/README.md
Spec authoring, docs/specs/ pathing, or TC format: feature-spec-reference.md, spec-system-reference.md, spec-principles.md
Behavior/public-contract changes or spec-test-code sync: workflow-spec-test-code-cycle-reference.md plus the spec docs above
Derived spec indexes/ERDs/reimplementation guides: spec-system-reference.md and source Feature Specs under docs/specs/
Integration test implementation/review: integration-test-reference.md
E2E test implementation/review: e2e-test-reference.md
Code review/audit work: code-review-rules.md plus domain docs above based on changed files

Do not read all docs blindly. Start from docs-index-reference.md, then open only relevant files for the task.

[BLOCKING] Execute skill steps in declared order. NEVER skip, reorder, or merge steps without explicit user approval. [BLOCKING] Before each step or sub-skill call, update task tracking: set in_progress when step starts, set completed when step ends. [BLOCKING] Every completed/skipped step MUST include brief evidence or explicit skip reason. [BLOCKING] If Task tools are unavailable, create and maintain an equivalent step-by-step plan tracker with the same status transitions.

Quick Summary

Goal: Block shipping an unproven fix — build a code proof trace (like a debugger stack trace) for each fix change with confidence percentages, so every code change carries a file:line evidence chain and confidence score and a fix ships only when its correctness is proven (≥80%), not assumed.

Summary:

Runs AFTER $fix (never before) as the non-negotiable verification gate between $fix and $code-simplifier; build a stack-trace-style proof chain (symptom → trigger path → root cause → fix mechanism → why-correct → edge cases → side effects) per change, every arrow carrying file:line.
Score each change with the points rubric (root cause +25, fix mechanism +20, pattern precedent +15, framework +10, edge cases +5 each, side effects +10, no regressions +5) → ≥80% ship, 60-79% flag, <60% BLOCK; pattern precedent (1+ working example) and edge cases (error/null/concurrent) are required, not optional.
Map every fix part to a primary/contributing/latent root cause from the hypothesis matrix and prove ALL feeder paths are closed; finish with cross-change checks (interaction, completeness, regression, dependency, performance incl. paging/index DB protocol) and a final verdict.
After the verdict, resolve the active Goal Contract: map each proof trace to its success criterion, update the Goal Satisfaction matrix — a SHIP verdict does NOT close work while any required criterion remains FAIL.

Workflow:

Inventory — List every code change made by the fix (file:line, before/after)
Trace — For each change, build a proof chain tracing from symptom → root cause → fix
Score — Assign confidence percentage per change with evidence
Verify — Cross-check fix against edge cases and side effects
Verdict — Overall fix confidence and any remaining risks

Key Rules:

Every claim MUST ATTENTION have file:line evidence — no exceptions
Each change gets its OWN proof trace and confidence score
If ANY change scores below 80%, flag it and recommend additional investigation
Always run this proof-trace step after $fix — non-negotiable, never skip it

Frontend/UI Context (if applicable)

When this task involves frontend or UI changes,

Component patterns: docs/project-reference/frontend-patterns-reference.md
Styling/BEM guide: docs/project-reference/scss-styling-guide.md
Design system tokens: docs/project-reference/design-system/README.md

Be skeptical. Apply critical thinking, sequential thinking. Every claim needs traced proof, confidence percentages (Idea should be more than 80%).

Prove Fix

Post-fix verification skill building evidence-based proof chains for every code change. A code debugger's stack trace, but for proving WHY a fix is correct.

When to Use

After $fix in bugfix, hotfix, or any fix workflow
After applying code changes that fix a reported bug
Verify fix correctness before review/commit

When NOT to Use

Before fix applied (use $debug-investigate instead)
New feature verification (use $test instead)
Code quality review (use $code-review instead)

Step 1: Change Inventory

List ALL changes made by the fix. For each change, document:

CHANGE #N: [short description]
  File: [path/to/file.ext]
  Lines: [start-end]
  Before: [code snippet — the broken version]
  After:  [code snippet — the fixed version]
  Type:   [root-cause-fix | secondary-fix | defensive-fix | cleanup]

Change types:

root-cause-fix — Directly addresses the root cause of the bug
secondary-fix — Fixes a related issue discovered during investigation
defensive-fix — Prevents the same class of bug from recurring
cleanup — Removed dead code or simplified logic (no behavior change)

Step 2: Proof Trace (per change)

For EACH change, build a stack-trace-style proof chain — the core of the skill.

Proof Trace Format

PROOF TRACE — Change #N: [description]
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

SYMPTOM (what the user sees):
  → [Observable behavior, e.g., "UI doesn't refresh after assigning PIC"]

TRIGGER PATH (how the symptom occurs):
  1. [file:line] User action → [method/event]
  2. [file:line] → calls [method]
  3. [file:line] → dispatches [action/event]
  4. [file:line] → handler/effect [name]
  5. [file:line] ← BUG HERE: [exact broken behavior]

ROOT CAUSE (proven):
  → [One sentence: what exactly is wrong and why]
  → Evidence: [file:line] shows [specific code proving the bug]
  → Hypothesis matrix mapping: RC-[N] status [primary/contributing/latent] from [report path/section]
  → Feeder paths closed: [path names or "single path verified"]; unresolved paths: [none/list]

FIX MECHANISM (how the change fixes it):
  → [One sentence: what the fix does differently]
  → Before: [broken code path with file:line]
  → After:  [fixed code path with file:line]
  → Owning fix layer: [layer/component] — why this is the lowest shared owner

WHY THIS FIX IS CORRECT:
  → [Reasoning backed by code evidence]
  → Pattern precedent: [file:line] shows same pattern working elsewhere
  → Framework behavior: [file:line or doc reference] confirms expected behavior
  → Forward convergence: [origin/trigger] -> [corrected transformations] -> [observed final output no longer stale/wrong]

EDGE CASES CHECKED:
  → [edge case 1]: [verified/not-verified] — [evidence]
  → [edge case 2]: [verified/not-verified] — [evidence]

SIDE EFFECTS:
  → [None / List of potential side effects with evidence]

CONFIDENCE: [X%]
  Verified: [list of verified items]
  Not verified: [list of unverified items, if any]

Proof Trace Rules

Every arrow (→) MUST ATTENTION have a file:line reference — no exceptions
TRIGGER PATH must be traceable — someone should be able to follow it step-by-step in the code
Hypothesis matrix mapping is REQUIRED for bugfixes — every fix part maps to a primary/contributing/latent root cause or is flagged as unrelated scope
Feeder paths must be accounted for — prove the fix closes every path that can write the final observed state, or explicitly list remaining unverified paths
Pattern precedent is REQUIRED — find at least 1 working example of the same pattern elsewhere in the codebase
Edge cases MUST ATTENTION be enumerated — at minimum: error path, null/empty input, concurrent access
Side effects MUST ATTENTION be assessed — what else could this change affect?
Spec-loop evidence is REQUIRED for a complete proof (canonical: SYNC:spec-loop-discipline) — the proof MUST carry, with file:line: (a) the regression property TC guarding the fixed invariant — a universally-quantified property ("for ALL inputs in {domain}, {invariant} holds") + boundary counter-case, not just the reproduction example; (b) mutation-kill evidence for the fixed core-logic line — show a surviving mutant on that line is now killed (MUTATION-SCORE bar, not line-coverage %); (c) a Dual-Feedback entry — the spec rule restored/added AND the guarding test that feeds it back. A proof trace missing any of (a)/(b)/(c) is INCOMPLETE — cap its confidence below the 80% ship threshold until the spec-loop evidence is supplied.

Step 3: Confidence Scoring

Each change gets an individual confidence score:

| Score | Meaning | Action Required | | ----------- | --------------------------------------------------------------------------- | -------------------------------------------------- | | 95-100% | Full proof trace complete, all edge cases verified, pattern precedent found | Ship it | | 80-94% | Main proof trace complete, some edge cases unverified | Ship with caveats noted | | 60-79% | Proof trace partial, some links unverified | Flag to user — recommend additional investigation | | <60% | Insufficient evidence | BLOCK — do not proceed until evidence gathered |

Scoring Criteria

Award points for each verified item:

| Criterion | Points | Evidence Required | | ----------------------------------------- | ------- | ----------------------------- | | Root cause identified with file:line | +25 | Code reference | | Fix mechanism explained with before/after | +20 | Code diff | | Pattern precedent found in codebase | +15 | Working example at file:line | | Framework behavior confirmed | +10 | Framework source or docs | | Edge cases checked (per case) | +5 each | Verification result | | Side effects assessed | +10 | Impact analysis | | No regressions identified | +5 | Test results or code analysis |

Total possible: 100+ (normalize to percentage)

Step 4: Cross-Verification

After individual proof traces, perform cross-change verification:

Interaction check — Do the changes interact with each other? Could one change break another?
Completeness check — Does the combined fix address ALL reported symptoms?
Regression check — Could the combined changes introduce new bugs?
Dependency check — Are there other code paths that depend on the changed behavior?
Performance regression check — Does the fix introduce performance issues?

[IMPORTANT] Database Performance Protocol (MANDATORY):

Paging Required — ALL list/collection queries MUST ATTENTION use pagination. NEVER load all records into memory. Verify: no unbounded GetAll(), ToList(), or Find() without Skip/Take or cursor-based paging.

Index Required — ALL query filter fields, foreign keys, and sort columns MUST ATTENTION have database indexes configured. Verify: entity expressions match index field order, database collections have index management methods, migrations include indexes for WHERE/JOIN/ORDER BY columns.

Step 5: Final Verdict

Produce a summary verdict:

FIX VERIFICATION VERDICT
━━━━━━━━━━━━━━━━━━━━━━━

Overall Confidence: [X%]

Changes Summary:
  #1: [description] — [X%] ✅/⚠️/❌
  #2: [description] — [X%] ✅/⚠️/❌
  #N: [description] — [X%] ✅/⚠️/❌

Symbols: ✅ ≥80% (ship) | ⚠️ 60-79% (flag) | ❌ <60% (block)

Remaining Risks:
  - [risk 1]: [likelihood] × [impact] — [mitigation]
  - [risk 2]: [likelihood] × [impact] — [mitigation]

Verification Method:
  - [Manual testing required? Which scenarios?]
  - [Automated tests cover this? Which tests?]
  - [Additional monitoring needed post-deploy?]

Recommendation: [SHIP / SHIP WITH CAVEATS / INVESTIGATE FURTHER / BLOCK]

Goal Satisfaction update (MANDATORY after verdict): resolve the active Goal Contract per SYNC:goal-contract-satisfaction-loop and map each proof trace to the saved success criterion it satisfies. Append proof evidence and remaining gaps to the goal file's Iteration Log and update its Goal Satisfaction matrix (PASS/FAIL/BLOCKED per criterion). A SHIP recommendation does NOT close the work while any required goal criterion remains FAIL — route the validated gap into another bounded fix loop or escalate a blocker.

Example: Proof Trace for NgRx Effect Fix

PROOF TRACE — Change #1: Move catchError inside switchMap
━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━━

SYMPTOM:
  → UI doesn't refresh after assigning PIC, Job Opening, or changing stage

TRIGGER PATH:
  1. candidate-quick-card-v2.component.ts:445 — User clicks "Assign PIC"
  2. candidate-card.container.component.ts:892 — onPersonInChargeChange($event)
  3. candidate-card.effect.ts:275 — SavePersonInCharge effect
  4. candidate-card.effect.ts:284 — dispatches LoadCandidateDetailsAction
  5. candidate-card.effect.ts:48 ← EFFECT IS DEAD — never processes the action

ROOT CAUSE:
  → catchError at outer pipe level (effect.ts:64) causes effect completion on ANY error
  → Evidence: effect.ts:43-69 shows catchError OUTSIDE switchMap
  → Evidence: ngrx-effects.js:156-165 confirms defaultEffectsErrorHandler
     only catches errors, not completions

FIX MECHANISM:
  → Move catchError INSIDE switchMap so errors are caught per-request
  → Before: effect.ts:64 — catchError at outer pipe → effect COMPLETES → DEAD
  → After:  effect.ts:52 — catchError inside switchMap → inner obs completes → outer SURVIVES

WHY THIS FIX IS CORRECT:
  → RxJS: catchError inside switchMap catches per-emission, outer stream continues
  → Pattern precedent: effect.ts:120 (moveApplicationToNextState) uses same inner pattern
  → Framework: NgRx effects auto-resubscribe on ERROR but NOT on COMPLETION

EDGE CASES:
  → 403 Forbidden: verified — returns SetCandidateDetails with isAllowDisplayed=false
  → Network timeout: verified — returns EMPTY, effect survives
  → Multiple rapid requests: verified — switchMap cancels previous (unchanged)

SIDE EFFECTS:
  → None — same error handling logic, only scope changed

CONFIDENCE: 95%
  Verified: root cause, fix mechanism, pattern precedent, framework source, all edge cases
  Not verified: behavior under specific proxy/auth middleware errors (very unlikely)

Run python .claude/scripts/code_graph trace <file> --direction downstream --json to prove fix doesn't break downstream.

Graph Intelligence (RECOMMENDED if graph.db exists)

If .code-graph/graph.db exists, enhance analysis with structural queries:

Verify test coverage: python .claude/scripts/code_graph query tests_for <function> --json
Trace affected code paths: python .claude/scripts/code_graph query callers_of <function> --json
Batch analysis: python .claude/scripts/code_graph batch-query file1 file2 --json

Graph-Trace for Fix Verification

When graph DB is available, use trace to PROVE the fix doesn't break downstream consumers:

python .claude/scripts/code_graph trace <fixed-file> --direction downstream --json — verify all downstream consumers, event handlers, and bus message listeners are unaffected
python .claude/scripts/code_graph trace <fixed-file> --direction both --json — full context: what triggered the bug (upstream) + what the fix affects (downstream)
Include trace results as evidence in the proof chain

Integration with Other Skills

This skill is the mandatory verification gate between $fix and $code-simplifier in fix workflows.

Workflow position:

... → $fix → $prove-fix → $code-simplifier → $review-changes → ...

If proof trace reveals issues:

Score ≥80%: proceed to next step
Score 60-79%: ask user whether to proceed or investigate further
Score <60%: BLOCK — return to $debug-investigate or $fix step

<prove-context>$ARGUMENTS</prove-context>

Workflow Recommendation

MANDATORY IMPORTANT MUST ATTENTION — NO EXCEPTIONS: If you are NOT already in a workflow, you MUST ATTENTION use a direct user question to ask the user. Do NOT judge task complexity or decide this is "simple enough to skip" — the user decides whether to use a workflow, not you:

Activate workflow-bugfix workflow (Recommended) — scout → investigate → debug → plan → fix → prove-fix → review → test

Execute $prove-fix directly — run this skill standalone

Next Steps

MANDATORY IMPORTANT MUST ATTENTION — NO EXCEPTIONS after completing this skill, you MUST ATTENTION use a direct user question to present these options. Do NOT skip because the task seems "simple" or "obvious" — the user decides:

"$code-simplifier (Recommended)" — Clean up fix implementation
"$integration-test" — Generate/update regression integration tests
"$workflow-review-changes" — Review all changes before commit
"Skip, continue manually" — user decides

[IMPORTANT] Use task tracking to break ALL work into small tasks BEFORE starting — including tasks for each file read. This prevents context loss from long files. For simple tasks, AI MUST ATTENTION ask user whether to skip.

Prerequisites: MUST ATTENTION READ before executing:

docs/project-reference/domain-entities-reference.md — Domain entity catalog, relationships, cross-service sync (read when task involves business entities/models)

External Memory: For complex or lengthy work (research, analysis, scan, review), write intermediate findings and final results to a report file in plans/reports/ — prevents context loss and serves as deliverable.

Evidence Gate: MANDATORY IMPORTANT MUST ATTENTION — every claim, finding, and recommendation requires file:line proof or traced evidence with confidence percentage (>80% to act, <80% must verify first).

End-to-Start Debugger Trace — For non-trivial bugs, failed verification, regression fixes, behavior-changing code, or unclear code flow, start from the observed final state and walk backward before proposing a fix.

Frame 0: observed end state — Name the exact user-visible output, failing assertion, log line, persisted value, API response, rendered UI, or aggregate bucket. Record the reader/query/renderer that produced it with file:line evidence.

Walk backward one hop at a time — Trace final reader -> projection/cache/storage -> writer -> consumer/handler/job -> producer/caller -> original trigger. At every hop record: input, transformation, output, owner, and evidence.

Enumerate all feeder paths — Find every upstream producer/caller/event/job that can write into the final path, including retry, async, cache, background, and alternate UI/API paths. Mark each path verified, ruled out, or still unknown.

Build the hypothesis matrix — For each plausible cause, list evidence for, evidence against, how to reproduce/verify, blast radius, and status (primary, contributing, ruled out, latent). Do not fix until competing causes are explicitly resolved or bounded.

Choose the owning fix layer — Identify the invariant owner and the lowest shared point that protects all downstream consumers. A fix at the symptom site is rejected unless the symptom site owns the invariant.

Prove convergence forward — After choosing the fix, walk start -> end again and show how the corrected state reaches the observed final output. Map each root cause to a fix part and each fix part to a test/proof.

BLOCKED until: final state named · backward trace written · all feeder paths enumerated · hypothesis matrix completed · owning fix layer justified · forward convergence proof mapped to tests.

NEVER: Start at the first suspicious code path. Collapse multiple producers into one "flow". Treat duplicate symptoms as duplicate records without proving the read model. Skip ruled-out hypotheses.

UI System Context — For ANY task touching .ts, .html, .scss, or .css files:

MUST ATTENTION READ before implementing:

docs/project-reference/frontend-patterns-reference.md — component base classes, stores, forms

docs/project-reference/scss-styling-guide.md — BEM methodology, SCSS variables, mixins, responsive

docs/project-reference/design-system/README.md — design tokens, component inventory, icons

Reference docs/project-config.json for project-specific paths.

Graph-Assisted Investigation — MANDATORY when .code-graph/graph.db exists.

HARD-GATE: MUST ATTENTION run at least ONE graph command on key files before concluding any investigation.

Pattern: Grep finds files → trace --direction both reveals full system flow → Grep verifies details

| Task | Minimum Graph Action | | ------------------- | -------------------------------------------- | | Investigation/Scout | trace --direction both on 2-3 entry files | | Fix/Debug | callers_of on buggy function + tests_for | | Feature/Enhancement | connections on files to be modified | | Code Review | tests_for on changed functions | | Blast Radius | trace --direction downstream |

CLI: python .claude/scripts/code_graph {command} --json. Use --node-mode file first (10-30x less noise), then --node-mode function for detail.

Critical Thinking Mindset — Apply critical thinking, sequential thinking. Every claim needs traced proof, confidence >80% to act. Anti-hallucination: Never present guess as fact — cite sources for every claim, admit uncertainty freely, self-check output for errors, cross-reference independently, stay skeptical of own confidence — certainty without evidence root of all hallucination.

Understand Code First — HARD-GATE: Do NOT write, plan, or fix until you READ existing code.

Search 3+ similar patterns (grep/glob) — cite file:line evidence

Read existing files in target area — understand structure, base classes, conventions

Run python .claude/scripts/code_graph trace <file> --direction both --json when .code-graph/graph.db exists

Map dependencies via connections or callers_of — know what depends on your target

Write investigation to .ai/workspace/analysis/ for non-trivial tasks (3+ files)

Re-read analysis file before implementing — never work from memory alone. — why: long context drifts from the file; the file is ground truth

NEVER invent new patterns when existing ones work — match exactly or document deviation. — why: divergent patterns fragment the codebase and slow every future reader

BLOCKED until: - [ ] Read target files - [ ] Grep 3+ patterns - [ ] Graph trace (if graph.db exists) - [ ] Assumptions verified with evidence

Fix-Layer Accountability — NEVER fix at the crash site. Trace the full flow, fix at the owning layer.

AI default behavior: see error at Place A → fix Place A. This is WRONG. The crash site is a SYMPTOM, not the cause.

MANDATORY before ANY fix:

Trace full data flow — Map the complete path from data origin to crash site across ALL layers (storage → backend → API → frontend → UI). Identify where the bad state ENTERS, not where it CRASHES.

Identify the invariant owner — Which layer's contract guarantees this value is valid? That layer is responsible. Fix at the LOWEST layer that owns the invariant — not the highest layer that consumes it.

One fix, maximum protection — Ask: "If I fix here, does it protect ALL downstream consumers with ONE change?" If fix requires touching 3+ files with defensive checks, you are at the wrong layer — go lower.

Verify no bypass paths — Confirm all data flows through the fix point. Check for: direct construction skipping factories, clone/spread without re-validation, raw data not wrapped in domain models, mutations outside the model layer.

BLOCKED until: - [ ] Full data flow traced (origin → crash) - [ ] Invariant owner identified with file:line evidence - [ ] All access sites audited (grep count) - [ ] Fix layer justified (lowest layer that protects most consumers)

Anti-patterns (REJECT these):

"Fix it where it crashes" — Crash site ≠ cause site. Trace upstream.

"Add defensive checks at every consumer" — Scattered defense = wrong layer. One authoritative fix > many scattered guards.

"Both fix is safer" — Pick ONE authoritative layer. Redundant checks across layers send mixed signals about who owns the invariant.

Source/test drift check. For coding, fix, debug, investigation, test, or review work: when source behavior changes, inspect affected unit/integration/E2E tests and decide from evidence whether tests should change to match intended behavior or the source change is an unintended bug to fix. Do not write tests for migration code; schema/data migrations are one-time execution paths, not core application logic.

AI Mistake Prevention — Failure modes to avoid on every task:

Re-read files after context changes. Context compaction, resume, or long-running work can make memory stale; verify current files before acting. Verify generated content against source evidence. AI hallucinates APIs, names, claims, and document facts. Check the relevant source before documenting or referencing. Check downstream references before deleting or renaming. Removing an artifact can stale docs, generated mirrors, configs, and callers; map references first. Trace the full impact chain after edits. Changing a definition can miss derived outputs and consumers. Follow the affected chain before declaring done. Verify ALL affected outputs, not just the first. One green check is not all green checks; validate every output surface the change can affect. Assume existing values are intentional — ask WHY before changing. Before changing a constant, limit, flag, wording, or pattern, read nearby context and history. Surface ambiguity before acting — don't pick silently. Multiple valid interpretations require an explicit question or stated assumption with risk. Keep shared guidance role-relevant. Universal guidance must help every receiving skill or agent; code-specific obligations belong only in code-specific protocols.

IMPORTANT MUST ATTENTION search 3+ existing patterns and read code BEFORE any modification. Run graph trace when graph.db exists.

MANDATORY IMPORTANT MUST ATTENTION cite file:line evidence for every claim. Confidence >80% to act, <60% = do NOT recommend.

IMPORTANT MUST ATTENTION read frontend-patterns-reference, scss-styling-guide, design-system/README before any UI change.

IMPORTANT MUST ATTENTION run at least ONE graph command on key files when graph.db exists. Pattern: grep → graph trace → grep verify.

IMPORTANT MUST ATTENTION trace full data flow and fix at the owning layer, not the crash site. Audit all access sites before adding ?..

MUST ATTENTION apply critical + sequential thinking — every claim needs appropriate traced evidence (file:line for repo/code claims; source URL or artifact section for research, product, content, and docs claims); confidence >80% to act, <60% DO NOT recommend. Anti-hallucination: never present guess as fact, admit uncertainty freely, cross-reference independently, stay skeptical of own confidence.

MUST ATTENTION apply AI mistake prevention — verify generated content against evidence, trace downstream references before deleting or renaming, verify all affected outputs, re-read files after context loss, and surface ambiguity before acting.

IMPORTANT MUST ATTENTION debugger trace gate: for non-trivial bug/fix/investigation/review work, start at the observed final output and trace backward through reader -> storage/projection -> writer -> consumer/job -> producer/trigger. Enumerate all feeder paths and hypotheses before fixing. BLOCKED until trace, hypothesis matrix, owning fix layer, and forward convergence proof exist.

MANDATORY Resolve the active Goal Contract BEFORE work (active plan goal.md → plans/goals/{YYMMDD-HHmm}-{slug}/goal.md → create from current request) and read saved success criteria before editing.
MANDATORY Append iteration evidence after execution; emit a Goal Satisfaction matrix (PASS/FAIL/BLOCKED) before reporting PASS; loop on validated FAIL; escalate repeated no-progress or blockers. NEVER store secrets in goal files.

Prompt-Enhance Closing Anchors

IMPORTANT MUST ATTENTION follow declared step order for this skill; NEVER skip, reorder, or merge steps without explicit user approval IMPORTANT MUST ATTENTION for every step/sub-skill call: set in_progress before execution, set completed after execution IMPORTANT MUST ATTENTION every skipped step MUST include explicit reason; every completed step MUST include concise evidence IMPORTANT MUST ATTENTION if Task tools unavailable, maintain an equivalent step-by-step plan tracker with synchronized statuses

Closing Reminders

IMPORTANT MUST ATTENTION Goal: block shipping an unproven fix — build a file:line proof trace (debugger-style: symptom → root cause → fix mechanism → why-correct → edge cases → side effects) and confidence score per change, so a fix ships only when correctness is PROVEN (≥80%), never assumed.

Protocols in force (concise digest of the SYNC/shared blocks this skill carries):

End-to-Start Trace: start at observed output, trace backward, hypothesis matrix before fixing.
UI System Context: read frontend-patterns, scss-styling, design-system before any UI change.
Graph Investigation: run ≥1 graph command on key files when graph.db exists.
Critical Thinking: traced file:line proof per claim; confidence >80% to act.
Understand Code First: grep 3+ patterns and read code before any modification.
Fix-Layer Accountability: trace full data flow, fix at owning layer not crash site.
Source/Test Drift: when source behavior changes, reconcile affected unit/integration/E2E tests from evidence.
AI Mistake Prevention: verify generated content against evidence, trace downstream references, verify all affected outputs, re-read after context loss, surface ambiguity.

IMPORTANT MUST ATTENTION every arrow (→) in every proof trace MUST ATTENTION carry a file:line reference — no exceptions; a claim without traced evidence is speculation and BLOCKS the ship. — why: a proof chain with an unproven link proves nothing. IMPORTANT MUST ATTENTION score each change with the points rubric (root cause +25, fix mechanism +20, pattern precedent +15, framework +10, edge cases +5 each, side effects +10, no regressions +5) → ≥80% ship · 60-79% flag to user · <60% BLOCK; NEVER ship a <80% change — route it back to $debug-investigate or $fix. — why: the threshold is the gate; bypassing it defeats the skill. IMPORTANT MUST ATTENTION run AFTER $fix, never before — this is the non-negotiable verification gate between $fix and $code-simplifier; NEVER skip it because the fix "looks obviously right". — why: obviousness is the exact illusion this gate exists to break.

Domain protocols this proof MUST NOT skip:

IMPORTANT MUST ATTENTION pattern precedent is REQUIRED, not optional — cite at least 1 working example of the same pattern at file:line elsewhere in the codebase; "no precedent found" caps confidence below 80%. — why: a fix with no precedent is an unproven invention. IMPORTANT MUST ATTENTION edge cases are REQUIRED — enumerate at minimum error path, null/empty input, concurrent access, each verified with evidence; assess side effects (what else this change touches). — why: the happy path passing is not the change being correct. IMPORTANT MUST ATTENTION spec-loop evidence is REQUIRED for a complete proof (SYNC:spec-loop-discipline) — (a) a regression PROPERTY TC (∀ inputs in {domain}, invariant holds) + boundary counter-case, (b) mutation-kill evidence on the fixed line (MUTATION-SCORE, not line-coverage %), (c) a Dual-Feedback entry (spec rule restored + guarding test); missing any one caps confidence below 80%. — why: line coverage and a single reproduction example let the original defect class survive. IMPORTANT MUST ATTENTION map every fix part to a primary/contributing/latent root cause from the hypothesis matrix and prove ALL feeder paths are closed; fix at the LOWEST shared owning layer, NEVER patch the crash/symptom site. — why: a symptom-site patch leaves the real cause writing bad state through other paths. IMPORTANT MUST ATTENTION for non-trivial bug/fix work, start at the observed final output and trace BACKWARD (reader → storage/projection → writer → consumer/job → producer/trigger) before proposing a fix; build the hypothesis matrix first. — why: starting at the first suspicious line fixes a symptom, not the cause. IMPORTANT MUST ATTENTION run the Database Performance Protocol in cross-verification — ALL list/collection queries paginated (no unbounded GetAll/ToList/Find), ALL filter/FK/sort columns indexed. — why: a correct fix that degrades query performance ships a new regression. IMPORTANT MUST ATTENTION when .code-graph/graph.db exists, run a trace --direction downstream on the fixed file to prove no downstream consumer breaks; include trace output as proof evidence. — why: a green local change can silently break a distant caller. IMPORTANT MUST ATTENTION resolve the active Goal Contract AFTER the verdict (SYNC:goal-contract-satisfaction-loop) — map each proof trace to its success criterion, emit the PASS/FAIL/BLOCKED matrix; a SHIP verdict does NOT close work while any required criterion remains FAIL. — why: a verified change is not a satisfied goal. MANDATORY IMPORTANT MUST ATTENTION break work into small task tracking todos BEFORE starting (one task per change/proof), keep exactly one in_progress, add a final review todo; on context loss the current task list first — resume, never duplicate.

Anti-Rationalization:

| Evasion | Rebuttal | | ------------------------------------------ | ------------------------------------------------------------------------------------------------ | | "Fix is obviously correct, skip the proof" | Obviousness is the illusion this gate breaks. Build the trace anyway — every arrow file:line. | | "No precedent, but it's clearly right" | No precedent caps confidence <80%. Find a working example or mark the change unproven. | | "Reproduction test passes, that's enough" | A single example ≠ a property. Add the universally-quantified regression TC + boundary case. | | "Line coverage is green" | Coverage ≠ kill. Show the surviving mutant on the fixed line is now killed (mutation score). | | "I'll just patch where it crashes" | Crash site ≠ cause site. Trace backward, fix the lowest invariant owner that protects all paths. | | "Score is 75%, close enough to ship" | <80% BLOCKS. Flag to user or return to $debug-investigate — never round a confidence up. |

IMPORTANT MUST ATTENTION run AFTER $fix only; every arrow carries file:line; score ≥80% to ship, <60% BLOCK — these three are the gate. (primacy-recency echo of the top three)

[TASK-PLANNING] Before acting, analyze task scope and break it into small task tracking todos systematically before starting — this is very important.

Hookless Prompt Protocol Mirror (Auto-Synced)

Source: .claude/.ck.json + .claude/skills/shared/sync-inline-versions.md (:full blocks) + .claude/scripts/lib/hookless-prompt-protocol.cjs

[WORKFLOW-EXECUTION-PROTOCOL] [BLOCKING] Workflow Execution Protocol — MANDATORY IMPORTANT MUST CRITICAL. Do not skip for any reason.

Generic portability boundary: Reusable skills and protocol text stay project-neutral; project-specific conventions are discovered from docs/project-config.json and docs/project-reference/. Apply shared AI-SDD from shared/sdd-artifact-contract.md. Read docs/project-config.json and docs/project-reference/docs-index-reference.md, then open the project reference docs named there. For spec, test-case, behavior-change, public-contract, or docs/specs/ work, route through the local spec docs named by the docs index: feature-spec-reference.md, spec-system-reference.md, spec-principles.md, and workflow-spec-test-code-cycle-reference.md when specs/tests/code must stay synchronized. If either file or a required reference doc is missing or stale, auto-run $project-init (or the narrow lower-level route such as $project-config, $docs-init, $scan-all, or $scan --target=<key>) before ordinary project-specific work. Any supported AI tool may execute when this shared context and local docs are available.

DETECT: If the prompt starts with an explicit slash skill/workflow command, execute it directly. Otherwise match the prompt against the workflow catalog and skill list.
ANALYZE: Choose the best option: execute directly, invoke a skill, activate a standard workflow, or compose a custom step combination.
AUTO-SELECT: Pick the best option yourself. Do not ask the user to choose between direct execution, skill, standard workflow, or custom workflow.
ACTIVATE: For a selected workflow, call $start-workflow <workflowId>; for a selected skill, invoke that skill; for a custom workflow, sequence custom steps directly; for direct execution, proceed with the task.
CREATE TASKS: task tracking for ALL workflow/skill/custom steps before execution when the selected path has multiple steps.
EXECUTE: Advance per the Workflow Step Advancement & Parallel Phases rule in your context instructions — model-driven; a sub-agent completion advances a step identically to an inline call; a parallel-phase group is an all-return barrier (advance only after ALL members return, never serialize it)

Shared AI-SDD Protocol Markers

Source: .claude/skills/shared/sync-inline-versions.md

SYNC:ai-sdd-artifact-contract

AI-SDD Artifact Contract — Shared spec-driven development rules stay portable and source-owned.

Keep reusable AI-SDD principles in .claude; put repository-specific paths, commands, owners, products, and formats in project config/reference docs.

Preserve cycle: spec -> plan -> tasks -> implement -> verify -> update spec/docs.

Trace every requirement or invariant through decision, task, TC/test, source evidence, and docs/spec update.

Treat code-to-spec extraction as reference-only until accepted by the canonical spec owner.

Any supported AI tool may plan, implement, review, or verify with synced context; using multiple tools is optional.

Update .claude source first, then sync generated mirrors; do not manually edit .agents, .codex, or AGENTS.md. — why: mirrors are generated artifacts; hand-edits are overwritten on the next sync

If docs/project-config.json, root instruction files, or a required project-reference doc is missing or stale, auto-run $project-init or the narrow lower-level route before ordinary project-specific work.

Active reference: shared/sdd-artifact-contract.md in the active skills root.

SYNC:ai-sdd-artifact-contract:reminder

MANDATORY Apply shared/sdd-artifact-contract.md; keep reusable AI-SDD in .claude and local rules in project docs.
MANDATORY Code-to-spec extraction is reference-only until canonical acceptance; any supported AI tool may execute with synced context.
MANDATORY Update .claude source before syncing generated mirrors; do not manually edit .agents, .codex, or AGENTS.md.
MANDATORY Missing or stale project config, root instruction files, or required reference docs route project-specific work through $project-init or the narrow setup route automatically. [TASK-PLANNING] [MANDATORY] BEFORE executing any workflow or skill step, create/update task tracking for all planned steps, then keep it synchronized as each step starts/completes.

[LESSON-LEARNED-REMINDER] [BLOCKING] Task Planning & Continuous Improvement — MANDATORY. Do not skip.

Break work into small tasks (task tracking) before starting. Add final task: "Analyze AI mistakes & lessons learned".

Extract lessons — ROOT CAUSE ONLY, not symptom fixes:

Name the FAILURE MODE (reasoning/assumption failure), not symptom — "assumed API existed without reading source" not "used wrong enum value".
Generality test: does this failure mode apply to ≥3 contexts/codebases? If not, abstract one level up.
Write as a universal rule — strip project-specific names/paths/classes. Useful on any codebase.
Consolidate: multiple mistakes sharing one failure mode → ONE lesson.
Recurrence gate: "Would this recur in future session WITHOUT this reminder?" — No → skip $learn.
Auto-fix gate: "Could $code-review/$code-simplifier/$security-review/$lint catch this?" — Yes → improve review skill instead.
BOTH gates pass → ask user to run $learn. [CRITICAL-THINKING-MINDSET] Apply critical thinking, sequential thinking. Every claim needs traced proof, confidence >80% to act. Anti-hallucination principle: Never present guess as fact — cite sources for every claim, admit uncertainty freely, self-check output for errors, cross-reference independently, stay skeptical of own confidence — certainty without evidence root of all hallucination. AI Attention principle (Primacy-Recency): Put the 3 most critical rules at both top and bottom of long prompts/protocols so instruction adherence survives long context windows. Goal-driven execution: Define success criteria first, loop until verified, and stop only when observable checks pass. Tests verify intent: Tests must protect business rules/invariants and fail when the protected intent breaks, not only mirror current behavior.

Common AI Mistake Prevention (System Lessons)

Re-read files after context compaction. Edit requires prior Read in same context; compaction wipes read state. Re-read before editing.
Grep for old terms after bulk replacements. AI over-trusts find/replace completeness. Grep full repo after bulk edits for missed refs in docs/configs/catalogs.
Check downstream references before deleting. Deletions cascade doc/code staleness. Map referencing files before removal.
After memory loss, check existing state before creating new. Compaction wipes prior-work memory. Query current state to resume — never blindly duplicate.
Verify AI-generated content against actual code. AI hallucinates APIs, class names, method signatures. Grep to confirm existence before documenting/referencing.
Trace full dependency chain after edits. Changing a definition misses downstream consumers. Trace the full chain.
When renaming, grep ALL consumer file types. Some file types silently ignore missing refs (no compile error). Search code, templates, configs, generated files.
Trace ALL code paths when verifying correctness. Code existing ≠ code executing. Trace early exits, error branches, conditional skips — not just happy path.
Update docs that embed canonical data when source changes. Docs inlining derived data (workflows, schemas, configs) go stale silently. Update all embedding docs alongside source.
Verify sub-agent results after context recovery. Background agents may finish while parent compacted — grep-verify output, don't trust assumed completion.
Cross-check full target list against sub-agent assignments. Parallel sub-agents by category miss boundary items. Reconcile union of assignments against target list before proceeding.
Sub-agents inherit knowledge only from their agent .md definition — use custom agent types, not built-in Explore. Tool adoption = permission + knowledge + enforcement (numbered workflow step).
Persist sub-agent findings incrementally, not as a final batch. Long sub-agents hit cutoffs before final write — findings lost. Instruct append-per-section to report file.
When debugging, ask "whose responsibility?" before fixing. Trace caller (wrong data) vs callee (wrong handling). Fix at responsible layer — never patch symptom site.
Grep ALL removed names after extraction/refactoring. Primary file "done" ≠ secondary files clean. Grep entire scope for every removed symbol before declaring complete.
Assume existing values are intentional — ask WHY before changing. Pattern-matching as "wrong" skips context. Before changing any constant/limit/flag: read comments, git blame, surrounding code.
Verify ALL affected outputs, not just the first. One build green ≠ all green. Multi-stack changes (backend/frontend/tests/docs) require verifying EVERY output.
Evaluate fit before copying a nearby pattern. Closest example ≠ matching preconditions — verify the new context shares the same constraints, base classes, scope, lifetime.
Holistic-first debugging — resist nearest-attention trap. Don't dive into first plausible cause. List EVERY precondition (config, env vars, paths, DB, endpoints, creds, versions, DI, data). Verify each against evidence (grep/query — not reasoning). Ask "what would falsify this?" — if nothing, it's not a hypothesis. Most expensive failure: going deeper in "obvious" layer while bug sits in layer never questioned.
Surgical changes — apply the diff test (context-aware). Two modes: (1) Bug fix → every line traces to the bug; no restyling; orphan cleanup only for imports YOUR changes made unused. (2) Review/enhancement → implement improvements AND announce as "Enhancement beyond main request: [what]". Never silently scope-creep. Diff test: "Would this line exist if I wasn't asked to do X?" — if no, delete or announce.
Surface ambiguity before coding — don't pick silently. Multiple valid interpretations → present each with effort: "[Request] could mean (1) [N h], (2) [N h]. Which matters?" List scope/format/volume/constraints assumptions first. If simpler path exists, say so. Never silently pick.
[MANDATORY FIRST ACTION] ALWAYS activate a suitable skill or workflow BEFORE responding. Match task against workflow catalog + skill list; invoke via skill invocation or $start-workflow <workflowId>. NEVER answer or write code before checking. Skip = protocol violation.
Why-Review adversarial mindset — apply when reviewing any plan, decision, or design. Default SKEPTIC not VALIDATOR: steel-man a rejected alternative, invert each stated reason ("what does it sacrifice?"), stress-test top 2-3 assumptions, run pre-mortem ("ships, fails in 3 months — what breaks?"), surface 1-2 alternatives author missed. Section presence ≠ quality; quality = causal reasoning + concrete mitigations + evidence, not "it's better" or "monitor closely".
Front-load report-write in sub-agent prompts for large reviews. Many-file sub-agents hit budget before final write — findings lost. Design prompts so: (1) report-write is first explicit deliverable, (2) append per-file/section (not batched), (3) scope bounded so reads don't exhaust budget. Truncated mid-sentence with no report file → spawn narrower scope, don't retry same prompt.
After context compaction, re-verify all prior phase outcomes before continuing. Summaries describe intent, not environment state (git index, filesystem, processes). On resume, FIRST audit: git status, re-read modified files, verify filesystem. Every "completed" claim is an untested hypothesis until evidence confirms.
OOM/memory: check row count before row size. Triage: (1) Unbounded query — no DB filter for trigger? Push filter to DB; eliminates OOM. (2) Large rows? Projection reduces proportionally. Row reduction > projection in ROI.
Keep domain concepts out of generic/shared/infrastructure layers. Reusable layer (shared library, framework, infra module) must reference NO consumer-specific domain concept — tenant/customer/product IDs, business entities, feature rules. Leak compiles + runs → passes review silently while coupling the "reusable" layer to one consumer. Keep shared type domain-free; push domain fields/logic down into the consumer via subclass/composition. — why: a layer coupled to one consumer's domain is no longer reusable.

Agent Skills: Prove Fix

Install this agent skill to your local

Skill Files