Skip to content
Agentmart
Submit a tool

retro

Review a session to propose improvements to skills, checklists or harness configuration.

Skill by krzysztofbury

IndexedMIT
Write a review

Chalk Labs

Need help deploying retro?

Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.

Book a call

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/krzysztofbury-retro

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Review a session to propose improvements to skills, checklists or harness configuration. Use for explicit /retro or a request to improve the working process, not an ordinary session summary or sign-off.

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/retro

What's inside

Skill name
retro
Description
Review a session to propose improvements to skills, checklists or harness configuration. Use for explicit /retro or a request to improve the working process, not an ordinary session summary or sign-off.
License
MIT
Instructions
59 lines in SKILL.md · view on GitHub

Files (1)

  • SKILL.md

SKILL.md (first lines)

# Retro

Turn supported lessons into approved changes. No change is a valid outcome.
Use the session evidence already available; disclose gaps after compaction
instead of reconstructing missing events or exporting the transcript.

## Configuration

- `LEDGER`: user-configured private log outside public/shared repositories.
  If unset, present the entry in chat and ask for a destination only when the
  user wants it saved. Do not silently create `RETRO.md` in the current repo.
- `MEMORY_INBOX` (optional): configured destination for personal-knowledge
  candidates, subject to its source/approval contract. Candidates are not facts.

## Workflow

1. Review the goal, outcome, useful practices, user corrections, failures and
   skill usage. Separate confirmed causes from hypotheses.
2. Classify supported findings as existing skill/checklist changes, new-skill
   candidates, harness configuration, or durable knowledge candidates.
3. Show a compact table: finding, proposed change, evidence and target file.
   Skip empty categories. Include removal or narrowing when warranted.
4. Apply skill/configuration edits only after approval of the specific proposal;
   approval may cover a subset. Route knowledge candidates through the configured
   inbox contract, or present them in chat when no inbox is configured.
5. Verify approved changes proportionately. Add a new entry at the top of the
   configured ledger, preserving all prior entries. Record actual changes,
   not planned ones. If no ledger is configured, return the entry in chat.

## Avoid instruction accumulation

Before proposing a new skill, look for the same failure pattern in the available
ledger. One occurrence supports a candidate; a second supports considering a
skill. Missing history is not proof of recurrence. Prefer improving an existing
workflow over adding a parallel one.

For an existing rule, ask what recurring failure it prevents and what current
evidence justifies its cost. Propose retirement, narrowing or consolidation if
the cause disappeared, the harness changed, or the rule duplicates or conflicts
with stronger guidance. Record the rule's scope, evidence and review date or
trigger. Retire it only after approval, preserving the decision history.

## Ledger entry

```markdown
## YYYY-MM-DD | harness | project
- Goal: reached / partly / no
- Helpful practice: evidence, or none
- Friction: confirmed cause or hypothesis
- Changes: actual files and retired rules, or none
- Candidates: pattern and occurrence count, or none
- Review: date or condition for revisiting changed rules
```

Keep secrets and personal information out of shared/public repositories.
Describe processes and evidence, not blame. A concise table and the resulting
diffs are enough; do not invent lessons to fill a template.

From krzysztofbury/llm-powerup, MIT.

More from krzysztofbury/llm-powerup

Signals

Stars
0
Forks
0
Open issues
0
Last commit (this folder)
Oct 7, 2026 (24 days ago)

From GitHub (fetched Oct 24, 2026).

Details

License
MIT
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from krzysztofbury

Done for you

Rather not set it up yourself?

Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.

See how we build AI agents →Read the case studies

Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.

launches shipped
50+
launches shipped
media placements
500+
media placements
founders on every account
2
founders on every account

Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block

Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.

Get retro running in your business

Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.

Goes to the Chalk Labs founders, not the newsletter. Privacy

Need help deploying retro?

CallGet help
Cite this page

retro skill. Agent Mart. https://agentmart.dev/skills/krzysztofbury-retro (last updated 2026-10-24)