agentleFS
cryptofish7cryptofish7/claude-skills/skills/reflect/SKILL.md

reflect

Skill1 starsChanged 9 months agoMITView on GitHub

Analyze the current conversation for learnings and persist approved insights. Triggers on "reflect", "retrospective", "session review", "what did we learn", "conversation review".

183 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use poteto-help, setup-pstack and 10 more.

Other files people use alongside this one.

Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
  • shiwenbin1617Skill

    poteto-help

    Guides users through pstack setup, /poteto-mode, and picking the skill, playbook, or principle for a task. Type /poteto-help with a question.

    Also in 16 projects that use this

  • bnemaSkill

    setup-pstack

    Configure which Pi models pstack uses per role. Detects available provider/model specs and writes pstack's user configuration. Use for /setup-pstack, "configure pstack models", or changing pstack model choices.

    Also in 27 projects that use this

  • eryzerzSkill

    make-bot-ui

    Use when building a custom UI (page, dashboard, buttons) that should wake a Grok Bot over a webhook, when the user must provide a webhook sender key, or when exposing that UI on Tailscale.

    Also in 20 projects that use this

  • jaydubya818Skill

    automate-me

    Use when the user wants working preferences captured in a reusable mode skill.

    Also in 43 projects that use this

  • jaydubya818Skill

    poteto-mode

    Use when non-trivial work should follow the smallest rigorous Jstack workflow.

    Also in 30 projects that use this

  • shiwenbin1617Skill

    principle-explain-the-number

    Apply before you measure, and before you trust, report, or act on a number you measured: a speedup, a regression, a throughput, a latency, or an eval result. Find what limits it, and rule out that it measured something other than the work you think.

    Also in 29 projects that use this

  • +6 more from jaydubya818

The file

We split this skill into the rules it gives an agent

4 rules. 0 are common, 2 are distinctive, and 2 are very specific to this project.

  • 0

    Common

    Most of the 173 edited versions we read keep these.

  • 2

    Distinctive

    This author's own choices. Worth a look.

  • 2

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Use AskUserQuestion with multi-select to let the user choose which findings to apply.

    line 68
  2. Very specific:

    Skill opportunities: Create a minimal stub at ~/.claude/skills/<name>/SKILL.md with frontmatter (name, description) and a # TODO section …

    line 73
  3. Distinctive:

    Check for duplicates: Read the existing "Mistakes to Avoid" section and project memory before proposing additions.

    line 82
  4. Distinctive:

    Memory candidates must be non-obvious: Don't propose things already captured in project docs, config files, or CLAUDE.md.

    line 84

“Common” is measured: kept by at least 6 in 10 of the 173 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Reflect

Analyze the current conversation session, extract actionable insights, and persist approved learnings to the appropriate destinations.

Workflow

Phase 1: Conversation Analysis

Scan the full conversation context for significant events:

  • Tasks completed — What was built, fixed, or changed
  • Errors encountered — Build failures, wrong commands, broken tests, misconfigurations
  • User corrections — Rejected approaches, redirections, explicit feedback
  • Decisions made — Architectural choices, trade-offs, tool selections
  • Patterns observed — Repeated workflows, friction points, workarounds used more than once

Filter ruthlessly. Only keep medium/high significance findings — things that would save time or prevent mistakes in future sessions.

Phase 2: Extract Insights

Classify each finding into one of five categories based on where it should be persisted:

CategoryDestination
Mistakes to avoidCLAUDE.md "Mistakes to Avoid" section (append)
Skill opportunities~/.claude/skills/<name>/SKILL.md (create stub)
Skill improvementsEdit existing skill's SKILL.md
Memory candidates~/.claude/projects/<path>/CLAUDE.md (append)
Process improvementsEdit CLAUDE.md workflow sections

For each finding, draft the exact content that would be written (the rule text, the stub, the edit, etc.).

Phase 3: Present Findings

Output a structured report:

## Session Reflection

**Session summary:** [1-2 sentence overview of what happened]
**Findings:** N total across M categories

### Mistakes to Avoid
1. [Finding title] — [Specific rule to add]

### Skill Opportunities
2. [Skill name] — [What it would do, why it's warranted]

### Skill Improvements
3. [Existing skill name] — [What to change and why]

### Memory Candidates
4. [Knowledge item] — [Why it's worth remembering]

### Process Improvements
5. [Workflow change] — [Before → After]

Skip empty categories. If the session is clean (no findings), report "Clean session — no actionable insights found." and stop.

Phase 4: User Decision

Use AskUserQuestion with multi-select to let the user choose which findings to apply. Present each finding as a numbered option.

Then execute only the approved actions:

  • Mistakes to avoid: Append to the "Mistakes to Avoid" section of the project's CLAUDE.md.
  • Skill opportunities: Create a minimal stub at ~/.claude/skills/<name>/SKILL.md with frontmatter (name, description) and a # TODO section describing the skill's purpose. User can flesh it out later with skill-creator.
  • Skill improvements: Show the diff before editing. Edit the existing skill's SKILL.md.
  • Memory candidates: Append to ~/.claude/projects/<path>/CLAUDE.md. Create the file if it doesn't exist.
  • Process improvements: Show the before/after for the affected CLAUDE.md section. Edit in place.

Guidelines

  • Be selective. 2 high-quality findings beats 10 marginal ones. If nothing significant happened, say so.
  • Rules must be specific and actionable. "Be careful with deployments" is noise. "Always verify on-chain bytecode after contract deployment — Foundry logs success during simulation before broadcast" is useful.
  • Check for duplicates. Read the existing "Mistakes to Avoid" section and project memory before proposing additions. Don't add what's already there.
  • Skill opportunities require evidence. Either 2+ occurrences of the same workflow in the session, or explicit user frustration with a repeated manual process.
  • Memory candidates must be non-obvious. Don't propose things already captured in project docs, config files, or CLAUDE.md.
  • Don't self-reference. This skill should never propose adding itself as a finding.
  • Respect scope. Project-specific knowledge goes to project memory. Universal lessons go to CLAUDE.md. Don't mix them.

Download now (ZIP)

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
What did you use it for? e.g. Reviewing pull requests in a Rails app

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.