specs
Interactively elicit decisions as a decision tree — frontier rounds, each with a recommended answer — and LOCK them into a durable formal spec, system-design…
Skill by IanLuo
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/ianluo-specs
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.
About
Use cases: documents, design, data, security
Reviews
Write a reviewNo 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
- Runs as
- Shell scripts
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/specs
What's inside
- Skill name
specs- Description
- Interactively elicit decisions as a decision tree — frontier rounds, each with a recommended answer — and LOCK them into a durable formal spec, system-design, or architecture doc. The spec is the complete requirements contract (problem, outcome, scope+flow, acceptance criteria, KPIs, NFRs, assumptions, data, rollback, security, verification — IEEE 29148 §9.6 shape). On re-entry, show the locked doc and re-elicit only what changed. Many tasks at once → batch mode — settle what they share in one pass, then per-task rungs. The lock stamps a contract (upstream/referrers) at the top. Use for "write a spec", "spec this task before building", "lock the requirements", "what are the NFRs / assumptions / data requirements", "update the architecture doc", "what's our system design", "re-open the spec". Do NOT use for repo indexing (init-context), one-shot research, visual design (design-task), or doc-vs-code review (review-task).
- License
- unknown
- Metadata
- domain: specification · audience: personal
- Instructions
- 135 lines in SKILL.md · view on GitHub
Before installing
- Ships executable scripts
Files (6)
- SKILL.md
- architecture.md · references/
- judging.md · references/
- spec.md · references/
- system-design.md · references/
- lock.sh · scripts/
The license isn't open source, so the instructions aren't shown here. Read SKILL.md on GitHub
More from IanLuo/skills
- 0.565score
herdr
IanLuo
Open and control herdr tabs/panes — the terminal workspace manager — to run claude (or pi, codex, gemini, …) in a separate VISIBLE terminal: create a tab…
SkillAFreshness: AClaude Code +9 - 0.565score
annotate
IanLuo
Render content as an annotatable HTML page the user can point at, then resolve pasted feedback back to the exact data it points at.
SkillAFreshness: AClaude Code +9 - 0.565score
canvas
IanLuo
Canvas exists so the agent's answers are understandable and the user's feedback keeps its meaning: an article-shaped live page on 127.0.0.1 — abstract, then…
SkillAFreshness: AClaude Code +9 - 0.565score
jev
IanLuo
Get an independent, scored second opinion about a piece of state from TypeSafe's System One model (Jev), called through cred.
SkillAFreshness: AClaude Code +9 - 0.565score
flagship
IanLuo
Manage projects, tasks, decisions, and knowledge via the flagship (fs) CLI — the persistent memory layer for agent orchestration.
SkillAFreshness: AClaude Code +9 - 0.559score
task-agent
IanLuo
Hand a task to a NEW agent working on its own git worktree of the same project, then collect and merge the result.
SkillAFreshness: AClaude Code +9 - 0.524score
skill-man
IanLuo
Create, revise, validate, audit, deploy, and sync personal skills in this repo, and decide what makes a good skill.
SkillBFreshness: BClaude Code +9 - 0.524score
librarian
IanLuo
Research questions with self-scoring subagents (one at default depth, targeted follow-up on gaps), curate verified results into a durable library at…
SkillBFreshness: BClaude Code +9 - 0.524score
dev-task
IanLuo
Start and run software development tasks with TDD/BDD, the house code-quality bar, implementation, and verification.
SkillBFreshness: BClaude Code +9 - 0.524score
credentials
IanLuo
Store and inject service credentials so agents can reach authenticated APIs without a secret value ever landing in the transcript.
SkillBFreshness: BClaude Code +9 - 0.524score
show-me
IanLuo
Help the user understand the current topic visually: pick the smallest view that makes the point clear — pseudocode for logic, call/component/file trees for…
SkillBFreshness: BClaude Code +9 - 0.519score
handoff
IanLuo
Carry a task across a context break.
SkillBFreshness: BClaude Code +9
Signals
- Stars
- 0
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Oct 6, 2026 (23 days ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- License unknown
- Repository
- github.com/IanLuo/skills
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from IanLuo
- 0.565score
jev
IanLuo
Get an independent, scored second opinion about a piece of state from TypeSafe's System One model (Jev), called through cred.
SkillAFreshness: AClaude Code +9 - 0.565score
canvas
IanLuo
Canvas exists so the agent's answers are understandable and the user's feedback keeps its meaning: an article-shaped live page on 127.0.0.1 — abstract, then…
SkillAFreshness: AClaude Code +9 - 0.565score
annotate
IanLuo
Render content as an annotatable HTML page the user can point at, then resolve pasted feedback back to the exact data it points at.
SkillAFreshness: AClaude Code +9 - 0.565score
flagship
IanLuo
Manage projects, tasks, decisions, and knowledge via the flagship (fs) CLI — the persistent memory layer for agent orchestration.
SkillAFreshness: AClaude Code +9 - 0.565score
herdr
IanLuo
Open and control herdr tabs/panes — the terminal workspace manager — to run claude (or pi, codex, gemini, …) in a separate VISIBLE terminal: create a tab…
SkillAFreshness: AClaude Code +9 - 0.559score
task-agent
IanLuo
Hand a task to a NEW agent working on its own git worktree of the same project, then collect and merge the result.
SkillAFreshness: AClaude Code +9
Cite this page
specs skill. Agent Mart. https://agentmart.dev/skills/ianluo-specs (last updated 2026-10-24)