agentleFS
farukzahrafarukzahra/agent-skills/skills/recap/SKILL.md

recap

Skill0 starsChanged 2 months agoMITView on GitHub

Summarize all user requests from the current AI chat session and produce a concise rich HTML recap (requested vs delivered). Use when the user invokes /recap or asks for a session summary, delivery recap, or handoff document.

  • Reads credentials

45 other projects use a version of this.

Adopted by

Most recently taken up by

Used with

Projects with it also use glossary, orchestrate and 7 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.
  • Microsoft AzureSkill

    glossary

    Short glossary of domain terms used by the team.

    Also in 3 projects that use this

  • varun29ankuSSkill

    orchestrate

    This skill should be used when the user asks to 'orchestrate a task', 'break down work into parallel agents', 'coordinate subtasks', 'run agents in parallel', or mentions 'multi-agent'. Decomposes complex tasks into tracked subtasks, dispatches parallel subagents, and coordinates until completion.

    Also in 4 projects that use this

  • sunnyclinches44Skill

    explain

    Explains code using real-world analogies and ASCII diagrams

    Also in 3 projects that use this

  • Microsoft AzureSkill

    build

    Build the TypeSpec Rust emitter from source. Run this after cloning the repo, after pulling changes, or when TypeScript source files have been modified.

    Also in 3 projects that use this

  • nicolasestremSkill

    brainstorm

    Structured idea generation and solution exploration with systematic analysis.

    Also in 3 projects that use this

  • BlockSkill

    plan

    Draft a TDD, PRD, or lightweight plan in the conversation without committing it.

    Also in 3 projects that use this

The file

We split this skill into the rules it gives an agent

12 rules. 0 are common, 7 are distinctive, and 5 are very specific to this project.

  • 0

    Common

    Most of the 39 edited versions we read keep these.

  • 7

    Distinctive

    This author's own choices. Worth a look.

  • 5

    Very specific

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

  • Common
  • Distinctive
  • Very specific
  1. Distinctive:

    Ensure local dev server: when the project rules require it (e.g.

    line 28
  2. Very specific:

    Read repo docs for start commands (TerapIA: npm run db:up + npm run dev → default http://localhost:3000).

    line 29
  3. Distinctive:

    If down, start it before finishing the recap.

    line 30
  4. Distinctive:

    Before generating HTML, apply anti-ui-slop and frontend-design:

    line 38
  5. Very specific:

    No generic slop: avoid cream #F4F1EA templates, default Inter/system-only stacks, and numbered cards without meaning.

    line 40
  6. Very specific:

    Ground in the product: when the repo has design tokens (e.g.

    line 42
  7. Very specific:

    Use a stable kebab-case data-topic-id (slug from title).

    line 58
  8. Distinctive:

    Keep copy short; max ~8–12 topic cards.

    line 79
  9. Distinctive:

    Use this dark token set when no project tokens exist:

    line 81
  10. Distinctive:

    Accurate: only claim delivery for work that was implemented or explicitly output.

    line 125
  11. Very specific:

    No secrets: never include tokens, passwords, or .env values.

    line 126
  12. Distinctive:

    If the recap reveals unfinished work, add a Pending section at the bottom (plain list).

    line 130

“Common” is measured: kept by at least 6 in 10 of the 39 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

/recap — Session delivery recap

Produce a single self-contained HTML file that recaps the conversation: what the user asked for (brief paraphrase) and what was actually delivered.

When to use

  • User invokes /recap
  • User asks for session summary, recap, handoff, or "what was asked vs delivered"
  • End of a long multi-topic agent session before closing

Workflow

  1. Read the full conversation (all user messages, chronological). Group related turns into topics; do not list every micro-message.
  2. For each topic, write at most two short blocks:
    • Requested — user intent in plain language (1–2 sentences). Use the project's UI language (Portuguese for TerapIA unless the project is English-only).
    • Delivered — concrete outcomes: files, routes, scripts, rules, tests; include paths when helpful.
  3. Skip failed or abandoned attempts unless the user should know about them.
  4. Generate HTML using the template below (inline CSS only, no external dependencies).
  5. Ensure local dev server when the project rules require it (e.g. AGENTS.md, user rules):
    • Read repo docs for start commands (TerapIA: npm run db:up + npm run dev → default http://localhost:3000).
    • If down, start it before finishing the recap.
    • Validate HTTP 200 on health or main route; record the effective URL (port may differ).
    • Include test login hints only if documented in the repo (never secrets from .env).
  6. Save to docs/session-recaps/YYYY-MM-DD-<slug>.html when the project has a docs/ folder; otherwise session-recap.html at workspace root or a path the user specifies.
  7. Reply with the file path, local URL, and a short markdown table summary (optional).

UI design (required)

Before generating HTML, apply anti-ui-slop and frontend-design:

  • No generic slop: avoid cream #F4F1EA templates, default Inter/system-only stacks, and numbered cards without meaning.
  • Dark mode by default for recap HTML (color-scheme: dark, meta color-scheme).
  • Ground in the product: when the repo has design tokens (e.g. TerapIA dashboard-theme-presets.ts → petroleo-atual), reuse primary, bg, ink, muted — do not invent a unrelated palette.
  • Typography: pair a restrained serif display (e.g. Literata) with a clean sans body (e.g. IBM Plex Sans); not Arial/Inter alone.
  • Hierarchy: eyebrow → title → local test panel (pill link grid + health dot) → clickable topic cards (mark as seen) → pending footer.
  • Accessibility: visible :focus-visible, prefers-reduced-motion, sufficient contrast on dark surfaces.

Topic cards — marcar como visto (required)

Each topic card is a conversation chunk the user can click to mark as read. State persists in localStorage (per file path + topic id).

  1. Markup — every topic:
    <article class="topic" data-topic-id="slug-curto-do-topico">
      <h2>Título do tópico</h2>
      ...
    </article>
    Use a stable kebab-case data-topic-id (slug from title). Required for persistence.
  2. Progress chip in header stats:
    <span class="chip chip--progress" data-recap-progress hidden>0/0 vistos</span>
  3. CSS — include styles from reference/topic-seen.css (or equivalent inline). Seen state: .topic--seen — lower opacity, desaturated, muted left bar, "Visto" pill top-right.
  4. JS — before </body>, inline the script from reference/topic-seen.js:
    • Click or Enter/Space toggles seen
    • role="button", aria-pressed, tabindex="0"
    • Storage key: recap-seen:{pathname}:{data-topic-id}
    • Updates [data-recap-progress] counter
  5. UX copy (optional hint below topics title): Clique em um card para marcar como visto.

Reference files (same folder as this skill): reference/topic-seen.css, reference/topic-seen.js.

HTML template

Keep copy short; max ~8–12 topic cards. Label text follows project locale (PT-BR: Pedido / Entregue / Pendências).

Use this dark token set when no project tokens exist:

TokenDefault
--bg-page#102223
--bg-card#16302f
--primary#2d6a6b
--secondary#7fb5b0
--ink#f2f5f4
--muted#9bb0ae

Structure (inline CSS + optional Google Fonts):

<!DOCTYPE html>
<html lang="pt-BR">
<head>
  <meta charset="utf-8" />
  <meta name="viewport" content="width=device-width, initial-scale=1" />
  <meta name="color-scheme" content="dark" />
  <title>Recap — {Project} — {Date}</title>
  <!-- Literata + IBM Plex Sans or project-appropriate pair -->
  <style>/* dark tokens, radial glow on body::before, card shadows */</style>
</head>
<body>
  <div class="wrap">
    <p class="eyebrow">{Project} · sessão agente</p>
    <header>...</header>
    <section class="local-env"><!-- pill link grid + health dot when dev server required --></section>
    <div class="topics">
      <p class="topics-hint">Clique em um card para marcar como visto.</p>
      <article class="topic" data-topic-id="example-topic">...</article>
    </div>
    <section class="pending">...</section>
    <footer>Generated by /recap</footer>
  </div>
  <script>/* inline topic-seen.js */</script>
</body>
</html>

Quality bar

  • Concise: each topic fits one card; no code dumps in HTML.
  • Accurate: only claim delivery for work that was implemented or explicitly output.
  • No secrets: never include tokens, passwords, or .env values.

Optional follow-up

If the recap reveals unfinished work, add a Pending section at the bottom (plain list).

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.