agentleFS
Sign inStart free
stevemarksavastevemarksava/superteam/skills/mirror/SKILL.md

mirror

Skill1 starsChanged 2 months agoMITView on GitHub

Opt-in self + peer reflection at task close — Superman reads the last 3 game tapes + reflection files, synthesises patterns, writes reflection to .squad/coach/mirror/. Trigger with /mirror after significant tasks.

6 other projects use a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

5 rules. 0 are common, 3 are distinctive, and 2 are very specific to this project.

  • 0

    Common

    Most of the 5 edited versions we read keep these.

  • 3

    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. Distinctive:

    Use /mirror at task close to get Coach's synthesis of recent performance.

    line 9
  2. Distinctive:

    After Vera sign-off, before the final message to Steve.

    line 14
  3. Very specific:

    Write the MIRROR output to .squad/coach/mirror/YYYY-MM-DD-[task-slug].md

    line 25
  4. Very specific:

    If a lesson candidate appears in 2 consecutive MIRRORs with the same substance, it is confirmed.

    line 52
  5. Distinctive:

    Never overwrite existing lessons

    line 55

“Common” is measured: kept by at least 6 in 10 of the 5 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

/mirror — Superman Reflection Synthesis

Use /mirror at task close to get Coach's synthesis of recent performance. Triggers a structured reflection — reads game tapes and Superman's self-review files, returns one actionable paragraph and writes the record.

When to use

After Vera sign-off, before the final message to Steve. Especially useful after:

  • A task where you deviated from the plan
  • A task where Vera failed first-pass
  • A long multi-agent sprint
  • Any task where something felt off

What Superman does

  1. Locate the last 3 game tapes in .squad/coach/game-tape/ — read them
  2. Locate the last 3 Superman self-reflection files — check .squad/telemetry/ for files matching *-superman-reflection.md, read the most recent 3
  3. Synthesise across all sources: what patterns repeat, what the team is learning, what held since last mirror
  4. Write the MIRROR output to .squad/coach/mirror/YYYY-MM-DD-[task-slug].md
  5. Append a one-line summary to ~/.claude/squad/agents/superman/memory/MEMORY.md under a ## Mirror Log section (create the section if absent)
  6. Report the synthesis to Steve in one paragraph — direct, honest, no softening

MIRROR output format

# MIRROR — [date] — [task-slug]

## Pattern Summary
[What repeated across the last 3 tapes — name the pattern, count its appearances]

## What Held
[What improved or stayed clean since last mirror — name it specifically]

## Score Delta
[Superman rubric score this task vs last mirror: +X improving / -X declining / = stable / N/A first mirror]

## One Instruction
[Exactly one sentence. Specific and actionable. Not encouragement.]

## Lesson Candidate
[Yes — draft: "[lesson text]"] OR [No — clean session]

Lesson trigger

If a lesson candidate appears in 2 consecutive MIRRORs with the same substance, it is confirmed. Coach writes it to ~/.claude/squad/agents/superman/lessons.md using the append protocol:

  • Prepend the new lesson at the TOP of the bullet list
  • Format: - **[YYYY-MM-DD] [topic]:** [lesson text] [source: mirror]
  • Never overwrite existing lessons
  • Maximum 20 lessons — when full, remove the 5 where compliance has been clean for 5+ consecutive scorable sessions

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.

Check your setup, free