Yeachan-Heo/autoresearch

<Use_When>

You already have a mission and evaluator from /deep-interview --autoresearch
You want persistent single-mission improvement with strict evaluation
You need durable experiment logs under .omc/autoresearch/
You want a supported path for periodic reruns via Claude Code native cron </Use_When>

<Do_Not_Use_When>

You need evaluator generation at runtime — use /deep-interview --autoresearch first
You need multiple missions orchestrated together — v1 forbids that
You want the deprecated omc autoresearch CLI flow — it is no longer authoritative </Do_Not_Use_When>

<Required_Artifacts> Canonical persistent storage lives under .omc/autoresearch/<mission-slug>/ and/or .omc/logs/autoresearch/<run-id>/.

Minimum required artifacts:

mission spec
evaluator script or command reference
per-iteration evaluation JSON
markdown decision logs

Recommended canonical shape:

.omc/autoresearch/<mission-slug>/
  mission.md
  evaluator.json
  runs/<run-id>/
    evaluations/
      iteration-0001.json
      iteration-0002.json
    decision-log.md

Reuse existing runtime artifacts when available rather than duplicating them unnecessarily. </Required_Artifacts>

<Cron_Integration> Claude Code native cron is a supported integration point for periodic mission enhancement. In v1, prefer documenting/configuring cron inputs over building a large scheduler UI.

If cron is used:

keep one mission per scheduled job
preserve the same mission/evaluator contract
append new run artifacts rather than overwriting prior experiments </Cron_Integration>

<Execution_Policy>

Do not hand execution back to omc autoresearch
Do not create multi-mission orchestration
Prefer reusing src/autoresearch/* runtime/schema helpers where they already match the stricter contract
Keep logs useful to humans, not only machines </Execution_Policy>

Files1

1 files · 1.0 KB

Select a file to preview

Overall Score

78/100

Grade

B

Good

Grades are signals, not a certification. Always review a skill yourself before use.

Safety

82

Quality

76

Clarity

75

Completeness

72

Summary

Autoresearch is a stateful skill for bounded, iterative improvement of a single mission. It manages an evaluator-driven loop that persists evaluation results and markdown decision logs under `.omc/autoresearch/`, continuing until explicit stop conditions (max-runtime, user cancellation, or terminal state) are met. The skill enforces a strict contract: single-mission-only in v1, structured JSON evaluator output with a required boolean `pass` field, and durable artifact tracking.

Detected Capabilities

file writestate managementjson parsingmarkdown generationmission/evaluator orchestrationruntime trackingdecision logging

Trigger Keywords

Phrases that agents use to match this skill to user intent.

autoresearch missionbounded iterative improvementevaluator-driven experimentmission enhancement loopdurable evaluation loggingmax-runtime iterationautoresearch cron

Use Cases

/deep-interview --autoresearch output processing
bounded iterative experiment improvement
durable mission evaluation logging
cron-based periodic mission enhancement
evaluator-driven experimentation with max-runtime guarantees

Quality Notes

Skill clearly defines scope boundaries: single-mission-only, strict contract around evaluator input format (JSON with `pass` boolean), and explicit stop conditions. This prevents unbounded loops and scope creep.
Strong artifact structure documented with clear canonical paths (.omc/autoresearch/, .omc/logs/autoresearch/). Supports both machine-readable (JSON evaluations) and human-readable (markdown decision logs) outputs.
Workflow is well-sequenced with distinct phases: confirm setup, track state, iterate, and terminate. Max-runtime is explicitly the primary stop condition.
Cron integration guidance is pragmatic and constrains behavior (one mission per job, append rather than overwrite).
Good guardrails: explicitly forbids multi-mission orchestration in v1, rejects deprecated `omc autoresearch` CLI, requires pre-existing mission/evaluator from `/deep-interview --autoresearch`.
Skill uses versioning (v1) to manage scope — helps signal future evolution without breaking current contract.
Minor clarity gap: 'mode/state is active for autoresearch' in workflow step 2 is somewhat abstract. Would benefit from concrete example of state structure or tracking mechanism.
No reference implementation or code examples provided — skill is interface/contract-focused but could be more concrete with pseudocode or example state snapshots.

Model: claude-haiku-4-5-20251001Analyzed: Oct 22, 2026

Reviews

Add this skill to your library to leave a review.

No reviews yet

Be the first to share your experience.

Version History

v1.1
Content updated
✦ AINo behavioral changes detected.
2026-10-22
LATEST
v1.0
2026-04-20
View This VersionInitial version

autoresearch