openclaw/0xcjl/cjl-autoresearch-cc

cjl-autoresearch-cc

Overview

Improve skills, prompts, articles, workflows, and systems via iterative mutation-testing.

Core principle: One small verifiable change per round. Large rewrites are unverifiable and will be reverted.

Workflow: small edits → test → score → keep improvements, discard regressions.

Inspired by Karpathy/autoresearch and 0xcjl/openclaw-autoresearch-pro.

Trigger Keywords

English: autoresearch

Chinese: 自动优化, 自动研究

Semantic Triggers (No Keywords Needed)

This skill activates when the user's intent matches, even without explicit keywords:

• User wants to improve any skill, prompt, article, workflow, or system
• User asks to polish, refine, enhance, or upgrade content
• User wants iterative testing and improvement
• User says '帮我改进一下这个prompt', 'optimize this'
• User says '迭代优化', '循环改进', '反复打磨'
• User asks '能不能更好', '如何提升质量'
• User uses "打磨", "精炼", "完善", "升级" in context of content improvement

Supported Optimization Targets

| Mode | Input | Example | |------|-------|---------| | Skill | Skill name or path | coding-standards, ~/.claude/skills/tdd-workflow | | Plugin | Path to a plugin directory | ~/.claude/plugins/everything-claude-code | | Prompt | A prompt text string | Inline or file path | | Article | An article/document text | Inline or file path | | Workflow | A process or workflow description | Inline or file path | | System | A mechanism or system design | Inline or file path |

Workflow

Step 1 — Identify Mode and Target

Before proceeding, confirm with user:

"Optimize [target] in [mode] mode? (yes/no)"

If no, ask for clarification. If yes, proceed to Step 2.

Parse the user's request to determine mode. Check for:

Keyword triggers:

• autoresearch [target] / 自动优化 [target] / 自动研究 [target]
• optimize [target] / improve [target] / 优化 [target] / 改进 [target]
• refine [target] / enhance [target] / 精炼 [target] / 增强 [target]

Semantic triggers (intent-based):

• User wants to improve any skill, prompt, article, workflow, or system
• User asks to polish, refine, enhance, or upgrade content
• User describes wanting iterative testing and improvement

Mode detection from intent:

| User Intent | Mode | |-------------|------| | Optimize a skill/SKILL.md file | Skill | | Optimize an agent configuration | Skill | | Improve a custom command | Skill | | Optimize a plugin | Plugin | | Improve hooks configuration | Plugin | | Improve a prompt text | Prompt | | Polish an article/document | Article | | Optimize a workflow/process | Workflow | | Improve a system mechanism | System |

For Skill/Plugin mode, resolve the path:

• Skill: ~/.claude/skills/<skill-name>/SKILL.md
• Plugin: ~/.claude/plugins/<plugin-name>/

If path doesn't exist, search in order: ~/.claude/skills/ → current dir → ask user.

Examples of semantic triggers (no keywords):

• "帮我优化一下这个skill" → Skill mode
• "这个prompt不够好，帮我改进" → Prompt mode
• "我想让这篇文章更通顺" → Article mode
• "优化一下部署流程" → Workflow mode

Step 2 — Generate Checklist (10 Questions)

Read the target content first. Then generate 10 diverse, specific yes/no checklist questions relevant to the content type:

For Skill/Plugin mode:

| # | Dimension | What to Check | |---|----------|---------------| | 1 | Description clarity | Is the description precise, actionable, and clear? Does it state what the skill does and when to use it? | | 2 | Trigger coverage | Does it cover main real-world use cases? | | 3 | Workflow structure | Are steps clearly sequenced and unambiguous? | | 4 | Error guidance | Does it handle error states and edge cases? | | 5 | Tool usage accuracy | Are tool names and parameters correct for Claude Code? | | 6 | Example quality | Do examples reflect real usage patterns? | | 7 | Conciseness | Is content free of redundant repetition? | | 8 | Freedom calibration | Is instruction specificity appropriate? | | 9 | Reference quality | Are references and links accurate? | | 10 | Completeness | Are all sections filled with real content? |

For Prompt mode:

| # | Dimension | What to Check | |---|----------|---------------| | 1 | Goal clarity | Does the prompt state a clear, specific goal? | | 2 | Role/tone | Is the desired role or tone specified? | | 3 | Input format | Is the input format clearly described? | | 4 | Output format | Is the expected output format specified? | | 5 | Constraints | Are key constraints and boundaries stated? | | 6 | Context sufficiency | Is enough context provided to avoid hallucination? | | 7 | Edge cases | Does it handle ambiguous or edge case inputs? | | 8 | Conciseness | Is it free of redundant or contradictory instructions? | | 9 | Actionability | Are instructions concrete and actionable vs. vague? | | 10 | Completeness | Are all necessary elements for the task present? |

For Article/Documentation mode:

| # | Dimension | What to Check | |---|----------|---------------| | 1 | Title quality | Does the title clearly convey the main value? Is it specific enough? | | 2 | Opening hook | Does the opening grab attention? Does it set clear expectations? | | 3 | Logical structure | Are ideas logically organized (not random)? | | 4 | Argument clarity | Are claims supported with evidence or reasoning? | | 5 | Conciseness | Is unnecessary padding or repetition removed? | | 6 | Transition flow | Do paragraphs/sections flow smoothly? | | 7 | Closing strength | Does the conclusion summarize and inspire action? | | 8 | Tone consistency | Is the tone consistent throughout? | | 9 | Readability | Is sentence/paragraph length varied appropriately? | | 10 | Audience match | Does language match the target audience level? |

For Workflow/System mode:

| # | Dimension | What to Check | |---|----------|---------------| | 1 | Goal clarity | Is the objective clearly stated? | | 2 | Step sequencing | Are steps in logical, efficient order? | | 3 | Completeness | Are all necessary steps present? | | 4 | Error handling | Are failure modes addressed (timeout, auth, network, resource exhaustion)? | | 5 | Edge cases | Are corner cases considered (empty input, large files)? | | 6 | Simplicity | Is the workflow/system as simple as possible? Can steps be combined or eliminated? | | 7 | Observability | Can progress/status be tracked? | | 8 | Reversibility | Can steps be undone if errors occur? | | 9 | Automation potential | Which steps could be automated? | | 10 | Maintainability | Is it easy to modify and extend? |

Present the 10 questions, numbered 1-10. Ask the user to select which ones to activate.

Rule: Must use at least 5 questions. Using fewer makes scoring unreliable.

After presenting, ask: "Ready to start the optimization loop? (yes/start)"

Step 3 — Prepare Test Cases

Test cases validate that mutations improve, not harm, the content. Generate realistic user scenarios.

• Skill/Plugin mode: Generate 3-5 realistic prompts a user would send when using the skill/plugin
• Prompt mode: Generate 3-5 test inputs that the prompt would process
• Article mode: Generate 3-5 ways the article might be read or consumed
• Workflow mode: Generate 3-5 scenarios the workflow would handle
• System mode: Generate 3-5 conditions the system would encounter

Store test cases in context — do not write to disk unless needed.

Step 4 — Run Autoresearch Loop

Tip: For mutation strategies, see Mutation Strategy Reference below.

Loop configuration:

• Rounds per batch: 30
• Max total rounds: 100
• Pause: After every 30 rounds, show summary and ask user to continue or stop
• Stop conditions:
  • User says stop
  • 100 rounds completed
  • Score reaches 100%
  • No improvement for 10 consecutive rounds

Per-round procedure:

Track progress: Round N/100 | Best: XX% | Last: +/-YY

Constraint: ONE mutation per round. Multiple changes = unverifiable = will be reverted.

1. Mutate: Make ONE small edit (see Mutation types)

2. Test: For each test case, simulate what output the content would produce

   Constraint: Be honest. If the output would disappoint a user, the mutation failed.

3. Score: Apply each active checklist question (0 or 1 per question). Score = (passed / total_questions) × 100

   Scoring scale:

   • 10/10 = 100% (perfect)
   • 7/10 = 70% (good)
   • 5/10 = 50% (minimum viable)

4. Decide: If new score ≥ best score → keep the mutation. If lower → revert

   Example: Best=85%, New=87% → Keep. Best=85%, New=83% → Revert.

   Trust the score. Don't rationalize a bad mutation.

5. Log: Round number, mutation type, score, keep/revert decision

Mutation types (pick ONE per round):

| Type | Name | When to Use | |------|------|-------------| | A | Add constraint | When content is too vague | | B | Strengthen coverage | When trigger cases are missing | | C | Add example | When steps are too abstract | | D | Tighten language | When words are soft ("try to") | | E | Error handling | When failure modes missing | | F | Remove redundancy | When content is verbose | | G | Improve transitions | When flow is choppy | | H | Expand thin section | When content is sparse | | I | Add cross-ref | When sections are isolated | | J | Adjust freedom | When balance is off |

Step 5 — Report Results (after each batch)

See Quick Reference below for output format examples.

After each batch (30 rounds):

Example:

Batch 1 (rounds 1-30):
  Best score: 85%
  Mutations kept: 23  |  Reverted: 7
  Most effective types: A, C, D

After full completion:

Optimized: [filename/path]
Score: XX% → YY% (+ZZ%)
Rounds: N (kept: K, reverted: R)
Top mutations: [type, type, type]
---
Final content:
[diff or inline]

Mutation Strategy Reference

High-impact, low-risk changes:

• Adding explicit constraints where the content is vague
• Expanding coverage to cover edge cases
• Adding concrete examples to abstract instructions
• Tightening soft language ("try to" → "must")

Avoid in one round:

• Large rewrites of entire sections
• Multiple unrelated changes at once
• Changing fundamental scope or purpose
• Formatting-only changes (no testable value)
• Adding content the user didn't request
• Removing more than 10% of content

Quick Reference

Keywords Reference

Auto-detect: autoresearch, 自动优化, 自动研究 Skill: autoresearch ~/.claude/skills/tdd Prompt: optimize this prompt: [text] Workflow: optimize the deployment workflow System: improve the error handling system

Semantic Triggers (No Keywords)

"帮我优化一下这个skill"           # → Skill mode
"这个prompt不太行"               # → Prompt mode
"我想让文章更通顺"               # → Article mode
"优化一下部署流程"               # → Workflow mode
"改进一下这个系统"               # → System mode
"improve this code review"         # → Prompt/Skill mode
"polish this documentation"       # → Article mode

Mode Detection

| Situation | Action | |-----------|--------| | Path detected | Skill/Plugin mode | | Keyword present | Keyword-specified mode | | Short text | Prompt mode | | Long document | Article mode | | Uncertain | Prompt mode (default) |

Edge cases: Empty → ask. Invalid path → fallback to ~/.claude/skills/. Ambiguous → ask.