Recap Doc

What it does

When your agent uses it

• : AI/Codex has implemented a feature and the user needs a guided walkthrough of what changed and why
• With blind-spot detection and anticipated questions

Example prompts

• “/recap-doc”

Requirements

• Pre-approved tools (allowed-tools): Read, Grep, Glob, Write, Bash(git:*), Bash(node:*), Skill

Workflow steps

5 steps, taken from the step headings in SKILL.md.

1. 1Scope Load
2. 2Evidence Collection (reuse tech-brief Stage 2)
3. 3Spec Cross-reference
4. 4AI Synthesis
5. 5Redaction + Write

What it can do on your machine

Read from SKILL.md and the folder at commit c9a2036. It shows what the files ask for, not the result of running them.

• Tool permissions

  Pre-approves these tools, so the agent can use them without asking each time:

  • Read
  • Grep
  • Glob
  • Write
  • Bash(git:*)
  • Bash(node:*)
  • Skill

  From allowed-tools in the SKILL.md frontmatter.

• Runs code

  Shell commands in SKILL.md call:

  • git

  From the folder's file list and the shell code blocks in SKILL.md.

• Network

  No URLs in SKILL.md. Its commands use git, which can reach the network depending on how they are called.

  From URLs in SKILL.md, links to its own repository left out.

• Credentials

  Names no API keys, tokens, secrets or passwords.

  From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Recap Doc loads about 2.8k tokens when it runs, and up to ~8.1k if it reads all its reference files. Until then it costs about 127 tokens; SKILL.md has 971 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it

~127

When it runs · the whole SKILL.md, loaded when a task matches

~2.8k

With references · SKILL.md plus every file in references/, read only if the agent opens them

~8.1k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from sd0xdev/sd0x-harness at commit c9a2036, republished under its MIT licence (© sd0xdev). 971 words, ~2,789 tokens.

name

recap-doc

description

Post-development recap document generator. Use when: AI/Codex has implemented a feature and the user needs a guided walkthrough of what changed and why, with blind-spot detection and anticipated questions. Not for: Q&A follow-up (use /recap-ask), technical share-out for teammates (use /tech-brief), or generic code explanation (use /codex-explain). Output: briefing-recap-<YYYY-MM-DD>.md with file-level walkthrough, design intents, spec drift, blind spots (mandatory), and anticipated questions.

allowed-tools

Read, Grep, Glob, Write, Bash(git:*), Bash(node:*), Skill

/recap-doc --scope <json-path-or-inline> [--focus <str>] [--depth brief|normal|deep] [--output <path>]

sequenceDiagram
    participant U as Caller (user or /post-dev-recap)
    participant D as /recap-doc
    participant S as scripts/detect-scope.js
    participant T as tech-brief-style collection
    participant CE as /codex-explain (Skill)
    participant SR as scripts/security-redact.js
    participant F as Output file

    U->>D: /recap-doc --scope <json> [--depth]
    D->>D: Phase 1: Load & validate ScopeReport
    D->>T: Phase 2: Collect git evidence for scope.files (reuse tech-brief Stage 2)
    D->>D: Phase 3: Cross-reference tech-spec (if feature_context.has_tech_spec)
    D->>CE: Phase 4a: Explain top-N changed files
    D->>D: Phase 4b: Synthesize sections + Blind Spots (Must) + Anticipated Questions
    D->>SR: Phase 5a: Scan output for high/medium-confidence secrets
    SR-->>D: Redacted or AbortError
    D->>F: Phase 5b: Write briefing-recap-<date>.md
    D-->>U: Emit output path + summary

# Recap: <feature-key or "session">
> **Scope source**: ...
> **Detected at**: ...
> **Focus**: ...
> **Confidence**: ...

## 1. Overview
## 2. Changed Files (table with file:line references)
## 3. Design Decisions
## 4. Spec vs Implementation Drift    (if has_tech_spec)
## 5. Blind Spots                     (FR-9 Must — any depth)
## 6. Anticipated Questions           (normal/deep only)
## 7. Evidence                         (commit SHAs, file:line index)

Input: /recap-doc --scope /tmp/scope.json --depth normal
Action: Load scope → collect git evidence for top-10 files → /codex-explain per file → synthesize sections including Blind Spots + 3+ Anticipated Questions → security-redact → write to <tmp>/sd0x-dev-flow-recap/briefing-recap-2026-04-17.md (ephemeral default; user opts in to commit via --output)

Input: /recap-doc --scope '{"version":1,"source":"uncommitted",...}' --focus "auth" --depth brief
Action: Parse inline JSON → filter to auth-related files → top-5 only → §5 Blind Spots top-3 only → omit §6 → write to <tmp>/sd0x-dev-flow-recap/briefing-recap-<YYYY-MM-DD>.md

Input: /recap-doc --scope scope.json --depth deep --output docs/features/<key>/briefing-recap-2026-04-17.md
Action: Load scope → top-15 with inline code snippets → full §5/§6 → redact → realpath-resolve target (must be inside repo root OR <tmp>) → write. Paths that escape both roots are rejected.

© sd0xdev, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 3 other files (references) in skills/recap-doc of sd0xdev/sd0x-harness.

• SKILL.md
• references/output-template.md
• references/prompt-template.md
• references/source-guide.md

Open the folder on GitHubat commit c9a2036