sd0xdev/sd0x-harness/tech-brief
Claude Code and agent skill by sd0xdev/sd0x-harness
- community
Checking that the SKILL.md file is still in the repository.
Technical briefing for developer sharing. Use when: sharing implementation findings with technical colleagues, post-development knowledge transfer, documenting what was built and why. Not for: PM/CTO summary (use project-brief), first-principles reasoning (use fp-brief), design-phase specs (use tech-spec). Output: 6-section technical brief with source provenance.
Install
npx vskill@latest install sd0xdev/sd0x-harness/tech-briefSKILL.md
The same SKILL.md was published earlier as sd0xdev/sd0x-harness/tech-brief.
Technical Briefing Skill
Trigger
- Keywords: tech brief, technical briefing, share with team, tech-brief, dev sharing, share findings, technical memo
When NOT to Use
| Scenario | Alternative | |
|---|---|---|
| PM/CTO executive summary (strip technical details) | /project-brief | |
| First-principles reasoning chain (why decisions were made) | /fp-brief | |
| Design-phase technical specification (before implementation) | /tech-spec | |
| Code explanation at function/file level | /codex-explain | |
| Simple document summary | Ask Claude directly |
Command Signature
| Flag | Default | Description | |
|---|---|---|---|
<feature-key> | Auto-detect | Feature key or docs path | |
--depth | normal | Output depth (brief/normal/deep) | |
--output | docs/features/<key>/5-tech-brief.md | Custom output path | |
--no-save | false | Print to stdout only |
Workflow
Phase 1: Context Resolution
- Parse
$ARGUMENTSfor feature-key, path, or flags - Resolve feature using
node scripts/resolve-feature.js [--feature <key>]— the wrapper, not the
CLI: it owns the failure payload. If scan_error !== false — including a payload with no such field, which is what a shell || echo '{}' fallback produces — the four source sets are unknown, not empty. Stop and take the ⚠️ Need Human exit; a non-null key is not evidence the sets are complete
- Load
doc_inventoryand the four source sets from resolver output - Validate paths (see Path Security)
Input Resolution Table
| Input Type | Example | Feature Resolution | Default Output Path | |
|---|---|---|---|---|
| Feature key | /tech-brief fp-brief | --feature fp-brief | docs/features/fp-brief/5-tech-brief.md | |
| Feature dir | /tech-brief docs/features/fp-brief/ | Extract key from path | docs/features/fp-brief/5-tech-brief.md | |
| Feature doc | /tech-brief docs/features/fp-brief/2-tech-spec.md | Extract key from parent | docs/features/fp-brief/5-tech-brief.md | |
| Non-feature | /tech-brief /tmp/notes.md | No feature context | Require --output | |
| No argument | /tech-brief | Auto-detect (5-level cascade) | Based on resolved feature |
Phase 2: Multi-Source Collection
Three-stage collection. See references/source-guide.md for detailed strategy.
Stage 1 — Document Collection: Read design_records (tech-spec, architecture, feasibility) for the why and current_authority for the what it does now. All optional — and a brief that presents a design record's claim as shipped behaviour is the failure mode to avoid; Stage 2's code evidence is what settles the difference.
Stage 2 — Code & Git Evidence: git log -20, git diff --stat, read top 5 changed source files (100 lines each) for file:line references.
Stage 3 — Request Selection: Glob all request docs (no status filter — completed features are the primary use case), max 3 by date desc. Extract ## References for threadIds and PR links.
Phase 3: Synthesis
- Build Source Provenance table (Section | Source Files | Confidence)
- For each output section, extract content from mapped sources (see
references/output-template.md) - Apply depth filter (section inclusion and detail level per depth matrix)
- Apply Evidence Insufficient Rule:
[Source unavailable — no <type> found for this feature]
Phase 4: Output
See Save Behavior for output path resolution.
Path Security
- Path normalization: Resolve
..and symlinks, verify repo boundary - Traversal rejection: Input containing
..is rejected - Output path:
--outputallows repo-external paths (e.g./tmp/), emit warning "writing outside repo" - Secret redaction: Before reading source docs, scan for high-confidence secret patterns (API keys, private keys) — high confidence: abort; medium confidence: mask
[REDACTED]
Depth Levels
| Level | Max Length | Description | |
|---|---|---|---|
| brief | ~500 words | Key points only — suitable for Slack sharing | |
| normal | ~1500 words | Full coverage with source citations | |
| deep | ~3000 words | Full coverage + code snippets + alternative comparison |
These are upper bounds, not targets. Source-thin features will produce shorter output.
Output
See references/output-template.md for full template and depth matrix.
Save Behavior
| Condition | Output Path | |
|---|---|---|
| Default (feature resolved) | docs/features/<key>/5-tech-brief.md | |
--output <path> | Specified path (warn if outside repo) | |
--no-save | stdout only, no file written | |
No feature + no --output | Gate: Need Human |
Preferred canonical name: 5-tech-brief.md. If 5- prefix is occupied, use next available number.
Verification
- [ ] Feature context resolved (5-level cascade or explicit)
- [ ] Path validation executed (no traversal, repo boundary)
- [ ] Secret redaction scan completed
- [ ] Source Provenance table present in output
- [ ] Each section cites source (or shows Missing Source marker)
- [ ] Output length within depth-level upper bound
- [ ] Evidence Insufficient markers used where source is thin
References
- Output template:
references/output-template.md - Source collection guide:
references/source-guide.md - Feature context resolution:
@skills/create-request/references/feature-context-resolution.md
Examples
Security
- Security scanPassed, 90 of 100Pattern checks on the skill's files.
- Model reviewPassed
Quality evaluation
- Task evaluationNot runNo evaluation yet. Run one in Skill Studio.
Source
- SKILL.md in the repositoryCheckingskills/tech-brief/SKILL.md
Popularity
- Repository stars
- 191
- Repository forks
- 26
- 7-day trend score
- 31
Stars and forks are for the whole GitHub repository, not this skill alone. Not refreshed yet.
Works with
Any agent vskill installs to: 56 targets in all. These 39 have a page on this site.
- Claude Code
- Cursor
- GitHub Copilot
- Windsurf
- Codex
- Gemini CLI
- Cline
- Amp
- Roo Code
- Goose
- Continue
- Kilo Code
- Junie
- Trae
- OpenHands
- Qwen Code
- AI Agent
- Tabnine Chat
- Sourcegraph Cody
- Amazon Q
- Firebase Genkit
- Sweep
- Devin
- Phind
- Mentat
- Claude Desktop
- Aider
- GPT Pilot
- Plandex
- bolt.new
- v0
- Replit Agent
- Anse
- TypingMind
- BetterChatGPT
- LibreChat
- Warp AI
- Pieces
- Fig AI
Embed badge
[](https://verified-skill.com/skills/sd0xdev/sd0x-harness/tech-brief)