Agent-Skills.md

Agent Skills: Bare Eval — Isolated Evaluation Calls

Run isolated eval and grading calls using CC 2.1.81 --bare mode. Constructs claude -p --bare invocations for skill evaluation, trigger testing, and LLM grading without plugin/hook interference. Use when running eval pipelines, grading skill outputs, benchmarking prompt quality, or testing trigger accuracy in isolation.

UncategorizedID: yonatangross/orchestkit/bare-eval

Repository

yonatangross/orchestkit
yonatangrossLicense: MIT
13314

Install this agent skill to your local

pnpm dlx add-skill https://github.com/yonatangross/orchestkit/tree/HEAD/src/skills/bare-eval

Skill Files

Browse the full folder contents for bare-eval.

Download Skill

Loading file tree…

src/skills/bare-eval/SKILL.md

Skill Metadata

Name
bare-eval
Description
"Run isolated eval and grading calls using CC 2.1.81 --bare mode. Constructs claude -p --bare invocations for skill evaluation, trigger testing, and LLM grading without plugin/hook interference. Use when running eval pipelines, grading skill outputs, benchmarking prompt quality, or testing trigger accuracy in isolation."

Bare Eval — Isolated Evaluation Calls

Run claude -p --bare for fast, clean eval/grading without plugin overhead.

CC 2.1.81 required. The --bare flag skips hooks, LSP, plugin sync, and skill directory walks.

When to Use

  • Grading skill outputs against assertions
  • Trigger classification (which skill matches a prompt)
  • Description optimization iterations
  • Any scripted -p call that doesn't need plugins

When NOT to Use

  • Testing skill routing (needs --plugin-dir)
  • Testing agent orchestration (needs full plugin context)
  • Interactive sessions

Prerequisites

# --bare requires ANTHROPIC_API_KEY (OAuth/keychain disabled)
export ANTHROPIC_API_KEY="sk-ant-..."

# Verify CC version
claude --version  # Must be >= 2.1.81

Quick Reference

| Call Type | Command Pattern | |-----------|----------------| | Grading | claude -p "$prompt" --bare --max-turns 1 --output-format text | | Trigger | claude -p "$prompt" --bare --json-schema "$schema" --output-format json | | Streaming grade | claude -p "$prompt" --bare --max-turns 1 --output-format stream-json | | Optimize | echo "$prompt" \| claude -p --bare --max-turns 1 --output-format text | | Force-skill | claude -p "$prompt" --bare --print --append-system-prompt "$content" | | @-file in prompt | claude -p "grade @fixtures/case-1.md against rubric" --bare (CC 2.1.113 Remote Control autocomplete) |

--output-format stream-json

Newline-delimited JSON events (one per token/tool-call) — lets a runner score partial output or abort early on a failing probe without waiting for the full response.

claude -p "$prompt" --bare --max-turns 1 --output-format stream-json \
  | while IFS= read -r line; do
      # line is a single JSON event; inspect $.type == "content_block_delta"
      jq -r 'select(.type == "content_block_delta") | .delta.text' <<< "$line"
    done

Use stream-json over json when:

  • grading long outputs and you want incremental scoring,
  • piping into another CLI step-by-step (e.g. ork:eval-runner),
  • you need per-token timing data alongside the content.

Invocation Patterns

Load detailed patterns and examples:

Read("${CLAUDE_SKILL_DIR}/references/invocation-patterns.md")

Grading Schemas

JSON schemas for structured eval output:

Read("${CLAUDE_SKILL_DIR}/references/grading-schemas.md")

Pipeline Integration

OrchestKit's eval scripts (npm run eval:skill) auto-detect bare mode:

# eval-common.sh detects ANTHROPIC_API_KEY → sets BARE_MODE=true
# Scripts add --bare to all non-plugin calls automatically

Bare calls: Trigger classification, force-skill, baseline, all grading. Never bare: run_with_skill (needs plugin context for routing tests).

CC 2.1.119: --print honors agent tools: / disallowedTools: (M122)

Before CC 2.1.119, --print mode ran with the full default tool set regardless of the agent's frontmatter tools: and disallowedTools:. Bare-eval grading was effectively ungated — graders could call any tool they wanted, even if the agent definition restricted them.

As of 2.1.119, --print enforces the agent's declared tool surface. Implications for eval design:

| Consequence | Action | |---|---| | Eval graders that relied on unrestricted tool access may now fail | Audit grader prompts for tools they actually need; whitelist explicitly via the agent's tools: frontmatter | | Eval results match interactive runs | Reproducibility improves — grading what the model can actually do, not what it could do in an unsandboxed --print | | --agent <name> also honors permissionMode in --print | Permission-gated tools (Bash, Edit) require either permissionMode: acceptEdits or explicit allowlists in the agent definition |

Migration test:

# Run an eval against an agent with a deliberately tight tools: list.
# Graders that previously called Read/Bash freely will now fail unless those
# tools are declared on the agent.
claude -p "$prompt" --bare --print --agent grader-test

If the grader fails with a "tool not permitted" error, add the required tool to the agent's tools: frontmatter and re-run.

CC 2.1.121: CLAUDE_CODE_FORK_SUBAGENT=1 for grader determinism (#1545)

Before CC 2.1.121, the env var only worked in interactive sessions. As of 2.1.121, non-interactive paths (claude -p, SDK) honor it too — each grader invocation gets a fresh forked subagent context.

The cross-eval state-leak problem this fixes:

Without forking, sequential claude -p --bare graders inherit harness state:

| Inherited | Symptom | |---|---| | memory MCP query cache | grader sees stale hit from previous run; same fixture grades differently | | .claude/chain/*.json on disk | grader for "implement" thinks "explore" already ran (file is from previous test) | | ToolSearch deferred-tool cache | first grader's MCP loads bleed into next grader's tool registry | | model picker pref | grader N inherits --model=opus from grader N-1 |

This produced ~5–10% retry rate and non-reproducible scores — the eval baseline drifted between runs, engineers chased phantom regressions.

Fix: tests/evals/scripts/lib/eval-common.sh exports CLAUDE_CODE_FORK_SUBAGENT=1, so every script that sources it (run-trigger-eval, run-quality-eval, run-agent-eval, optimize-description, etc.) gets forked graders automatically. The CI workflow .github/workflows/orchestkit-eval.yml also sets it at the workflow level. Older CC silently ignores the env var (no-op).

Determinism contract: running the same grader on the same fixture twice in a row produces the same score. Verified by tests/evals/scripts/test-grader-determinism.sh.

Performance

| Scenario | Without --bare | With --bare | Savings | |----------|---------------|-------------|---------| | Single grading call | ~3-5s startup | ~0.5-1s | 2-4x | | Trigger (per prompt) | ~3-5s | ~0.5-1s | 2-4x | | Full eval (50 calls) | ~150-250s overhead | ~25-50s | 3-5x |

Rules

Read("${CLAUDE_SKILL_DIR}/rules/_sections.md")

Troubleshooting

Read("${CLAUDE_SKILL_DIR}/references/troubleshooting.md")

Related

  • eval:skill npm script — unified skill evaluation runner
  • eval:trigger — trigger accuracy testing
  • eval:quality — A/B quality comparison
  • optimize-description.sh — iterative description improvement
  • Version compatibility: doctor/references/version-compatibility.md