meta-thinking
Skill3 starsChanged 6 months agoMITView on GitHub
Use this skill when the user needs structured planning, complex analysis, difficult tradeoff decisions, or troubleshooting under uncertainty and the agent should apply a compact meta-thinking loop with task classification, a meta-mode gate, plan extraction, critique, and concise trust signals. Prefer it when surfacing confidence, limitations, and context gaps will materially improve the answer. Do not use it for simple factual or one-step requests where extra framing would add friction without improving the result.
1 other project uses a version of this.
Adopted by
Most recently taken up by
- shreyas-makes/agent-skillstheir own version, first seen Jul 15, 2026
Other versions1 project uses a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 1
- shreyas-makes/agent-skills★ 2same name
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository18 other files it gives its agents.
- AGENTS.mdAGENTS.md
- agent-browserSkill
- ai-sdkSkill
- electron-scaffoldSkill
- electronSkill
- grill-meSkill
- ou-features-auditSkill
- ou-full-code-reviewSkill
- ouroboros-intent-e2eSkill
- ouroboros-test-plan-authorSkill
- prd-to-ticketsSkill
- team-implement-ticketsSkill
- ai-sdkSkill
- grill-meSkill
- ou-full-code-reviewSkill
- prd-to-ticketsSkill
- refactor-agents-mdSkill
- team-implement-ticketsSkill
The file
Show its rules and text
We split this skill into the rules it gives an agent
15 rules. About 0 are common, 15 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈15
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 10
Use this skill to apply the SecondOrder method to the agent's own work.
- Distinctive: line 14
Use this skill when the request benefits from one or more of:
- Distinctive: line 22
Do not use the full workflow for:
- Distinctive: line 37
Treat this as a compact control loop, not an excuse to dump internal reasoning.
- Distinctive: line 41
When meta mode is justified, bias toward a compact response structure that can include:
- Distinctive: line 51
Only surface fields that materially help the user.
- Distinctive: line 57
Before finalizing, check for:
- Distinctive: line 65
If the critique finds meaningful issues, tighten the answer and surface only the user-relevant residue.
- Distinctive: line 78
Do not reveal chain-of-thought or hidden scratch work.
- Distinctive: line 79
Do not force every answer into a verbose template.
- Distinctive: line 80
Do not use this as a generic "think harder" wrapper.
- Distinctive: line 81
Do not turn the task into a lesson on metacognition unless the user asks for that.
- Distinctive: line 85
Read references/second-order-method.md for the canonical method, terminology, and gating heuristics.
- Distinctive: line 86
Read references/output-patterns.md for compact response patterns by task type.
- Distinctive: line 87
Read references/examples.md for realistic trigger and non-trigger examples.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Meta Thinking
Use this skill to apply the SecondOrder method to the agent's own work. This is an execution skill, not a teaching curriculum.
Activation
Use this skill when the request benefits from one or more of:
- structured planning before answering
- tradeoff analysis or decision support
- troubleshooting with missing or unreliable context
- explicit handling of uncertainty, limitations, or context gaps
- reflective response shaping instead of a direct one-pass answer
Do not use the full workflow for:
- simple factual lookup
- short transformations or edits with obvious steps
- routine requests where standard tooling is sufficient and visible framing would add noise
Default Workflow
- Classify the task.
- Decide whether meta mode is warranted.
- Extract the goal, constraints, and response strategy.
- Produce a compact plan.
- Critique the draft for weak assumptions, limitations, and missing context.
- Return the answer with concise visible framing when justified.
Treat this as a compact control loop, not an excuse to dump internal reasoning.
Output Contract
When meta mode is justified, bias toward a compact response structure that can include:
goalconstraintsplanconfidencelimitationscontext_gapsresponse_strategywhen it helps the answer land cleanly
Only surface fields that materially help the user. Omit empty sections. Keep the framing short enough that the answer still feels direct.
For response shapes by task type, read references/output-patterns.md.
Critique Pass
Before finalizing, check for:
- missing constraints
- unsupported assumptions
- plan steps that are too broad or too brittle
- uncertainty that should be made explicit
- context gaps that, if resolved, would change the answer materially
If the critique finds meaningful issues, tighten the answer and surface only the user-relevant residue.
Stop Conditions
De-escalate to a direct answer when:
- the task is low complexity
- the plan adds no meaningful value
- the critique produces no user-relevant caution
- visible meta framing would be longer than the useful content
Anti-goals
- Do not reveal chain-of-thought or hidden scratch work.
- Do not force every answer into a verbose template.
- Do not use this as a generic "think harder" wrapper.
- Do not turn the task into a lesson on metacognition unless the user asks for that.
References
- Read references/second-order-method.md for the canonical method, terminology, and gating heuristics.
- Read references/output-patterns.md for compact response patterns by task type.
- Read references/examples.md for realistic trigger and non-trigger examples.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

