mcpField
D
debrief

debrief

github.com/paulrbergAgent workflows102113 installsUpdated today
Repository

About debrief

debrief is a task skill for turning evidence from the current task into an interactive HTML report or concise Markdown. It guides report creation, evidence checks, and inspection of the finished file.

Key features

  1. 01Interactive HTML report output
  2. 02Terse Markdown mode
  3. 03Topical slug generation
  4. 04Evidence-grounded content checks
  5. 05Output and overwrite safeguards

Use cases

  1. 01Recording verified findings and risks
  2. 02Reviewing code or document changes
  3. 03Sharing task conclusions as Markdown

Install

Copy the command and run it in a terminal.

$npx skills add https://github.com/paulrberg/agent-skills --skill debrief

After installing, type/debriefin Claude Code to confirm it loaded.

Bundled filesagents/openai.yamlscripts/prepare.sh

Details

From the project README

Debrief

Persist evidence from the current task as an opinionated interactive HTML report, or as terse Markdown with --md.

Output and Authority

  • Default: ./.ai/debriefs/<slug>/index.html using the installed playground skill.
  • --md: ./.ai/debriefs/<slug>/index.md, with no playground dependency.
  • Derive a short topical kebab-case slug when omitted.
  • Write only inside the selected debrief directory. If the output exists, stop and ask whether to overwrite or choose a new slug.
  • Use only evidence present in the current task: real findings, paths, verified locations, metrics, decisions, and unresolved risks. Do not invent filler.

Workflow

  1. Run the bundled preparer from the target repository:

    bash <skill-dir>/scripts/prepare.sh [--md] <slug>

    Use its MODE, PLAYGROUND_DIR, DEBRIEF_PATH, and EXISTS output. Relay dependency or slug errors and stop.

  2. For HTML, read the resolved playground SKILL.md and exactly one matching template: diff review, document critique, code map, concept map, data explorer, or design playground. Adapt its interaction model to the evidence; do not load every template.

  3. Build one self-contained file:

    • HTML keeps the playground's live controls, useful presets, evidence view, natural-language prompt output, and copy feedback. No external dependencies.
    • Markdown leads with the takeaway, then concise findings with severity/evidence/suggestion and actionable next steps. Use snippets only when exact text matters.

  4. Verify that every claim traces to the task transcript or tool evidence, the output contains no placeholders, and the file opens/renders. Then run open "$DEBRIEF_PATH" and report the absolute path.

Completion requires a non-placeholder debrief at the selected path, evidence-grounded content, successful rendered/file inspection, and explicit overwrite handling.

Comments

Markdown supported