Agent-Skills.md

Agent Skills: Octocode Research

Use when researching, debugging, reviewing, or changing code with evidence-first proof. Covers local and GitHub repos, npm packages, PRs, dead code, architecture, prior art, and iterative ReAct loops.

UncategorizedID: bgauryy/octocode-mcp/octocode-research

Repository

bgauryy/octocode-mcp
bgauryyLicense: MIT
88173

Install this agent skill to your local

pnpm dlx add-skill https://github.com/bgauryy/octocode/tree/HEAD/skills/octocode-research

Skill Files

Browse the full folder contents for octocode-research.

Download Skill

Loading file tree…

skills/octocode-research/SKILL.md

Skill Metadata

Name
octocode-research
Description
"Use when researching, debugging, reviewing, or changing code with evidence-first proof. Covers local and GitHub repos, npm packages, PRs, dead code, architecture, prior art, and iterative ReAct loops."

Octocode Research

Evidence-first technical research and code work: SCOPE -> SEARCH -> READ EXACT -> VALIDATE -> DECIDE/PATCH -> VERIFY.

Modes

Investigate behavior/root cause; Review PRs/diffs; Change after evidence; Map/Validate prior art or architecture; Loop when evidence shifts.

Rules

  1. State corpus, question, mode, and active/skipped surfaces in one line.
  2. Route by what you already hold; never force a fixed grep -> AST -> LSP path.
  3. For nontrivial code claims, read at least two of structure, stream, and connections.
  4. Keep a tiny ledger: claim -> evidence -> confidence -> next check.
  5. Ask before broad public contracts, deletes/renames, thin evidence, or 3+ unrelated problem spaces.
  6. For code edits, make the smallest scoped patch and report the checks that actually ran.

Reference Map

Before any task, read references/algorithm.md first; it owns routing, evidence grades, and failure signals.

  • references/workflows.md — when choosing a mode-specific workflow index.
  • references/code-research.md — when investigating, reviewing, refactoring, or changing code.
  • references/research-flow.md — when doing general research or idea validation.
  • references/workflow-local.md — when the local repo or installed dependency is source of truth.
  • references/workflow-external.md — when using GitHub, npm, PRs, commits, or remote repos.
  • references/workflow-debug.md — when debugging failures or proving root cause.
  • references/workflow-pr-review.md — when reviewing PRs, diffs, or merge safety.
  • references/workflow-pr-review-analysis.md — when PR review analysis and finding shape are needed.
  • references/workflow-pr-review-report.md — when writing PR/local review summaries or documents.
  • references/workflow-change.md — when implementing or refactoring after evidence.
  • references/github-landscape.md — when ranking GitHub repos or ecosystem prior art.
  • references/long-research.md — when a decision brief, audit trail, or contested question needs depth.
  • references/loop-mode.md — when evidence keeps shifting or verification fails repeatedly.
  • references/octocode.md — when command syntax, MCP transport, or raw tool schemas matter.

Scripts

  • scripts/eval-research.mjs — when changing this skill; run the matching eval case.

Output

Quick answer: Finding, Evidence, Confidence, Next. Decision/review: TL;DR, evidence, verdict, risks, file:line, verification, confidence, smallest safe fix.