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/bibutikoley-spec
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: design, devops, 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
Network hosts it talks to 1
codeload.github.com· found in install command
Files it touches
~/.claude/skills/spec
What's inside
- Skill name
spec- Description
- Stage 2 (Design) of the AI-native SDLC. Read an accepted intent.md and produce a requirements-and-design spec.md constrained by the organisation's policy skills (security, compliance, brand, UX), with flagged concerns first. Use when an intent is accepted, when asked for a spec/requirements/design from an intent, or on /sdlc:spec. Also used non-interactively by CI.
- License
- MIT
- Instructions
- 54 lines in SKILL.md · view on GitHub
Files (1)
- SKILL.md
SKILL.md (first lines)
# /sdlc:spec — requirements and design in one pass Read `sdlc-loop`. The product owner reviews this spec; they do not write it. ## Preconditions 1. Resolve the change: `$ARGUMENTS` is a change id, a path to `intent.md`, or empty (then list changes whose `intent.md` is `status: accepted` and has no `spec.md`, and ask which one). 2. `intent.md` must be `status: accepted`. If it is `draft`, stop and say the product owner must accept it first (merge). If the user explicitly overrides, proceed and write `override: intent not accepted` in the spec frontmatter. 3. Load every policy skill available: `security-policy`, `compliance-policy`, `brand-guidelines`, `ux-standards`, `secure-api-review`, plus any under the repo's `.claude/skills/`. Record which ones you applied in `skills_applied` (name@last-reviewed date from the skill header). 4. If a Figma URL is given or referenced in the intent and the Figma connector is available, read the frames with `get_design_context` / `get_screenshot` and treat them as the approved mock. 5. If the tracker holds the record (`sourceOfTruth: tracker|linked`), read the ticket first (`tracker-sync`). ## Produce `spec.md` Use `sdlc/changes/_template/spec.md`. The prompt you are executing is, in the course's words: > Read the intent.md and produce a requirements and design spec for integrating it into our existing codebase. > Apply the skills available to you so the spec conforms to our brand guidelines, security policies and UX > standards. Document the spec fully as spec.md, ready to hand to the engineering team. Describe clearly any > areas of concern, especially where you cannot satisfy contradicting policies. Rules: - **Flagged concerns come first.** Every place a policy conflicts with the intent, with another policy, or with the codebase gets a row: concern, policy/skill, owner who must decide, decision (blank until made). These are the points an analyst would have escalated. - Read the codebase (read-only) so the design names real modules, endpoints and conventions. Cite paths. - Requirements are numbered and testable and each traces to a line in `intent.md`. - Acceptance criteria are observable checks an engineer can turn into tests. - Carry every open question from `intent.md` forward as answered (where) or still open (who answers). - Set `intent_commit` to the SHA of the accepted intent.md. Inherit `risk`; raise it if the design reveals money movement, auth changes, PII, regulated data or infra. ## Commit and hand over - Branch `sdlc/spec/<change-id>`, commit `spec: <title>` with trailer `SDLC-Change: <id>`, PR titled `Spec: <title>` assigned to the product owner (from `owner`). In CI, label the PR `sdlc:spec`. - If the tracker is authoritative, write a summary + SHA back to the ticket. - Report: the path, the list of flagged concerns with owners, and the gate: **the product owner resolves each flagged concern with its policy owner, then accepts by merging with `status: accepted`. A tech lead must also approve if `risk: high`. Acceptance starts plan mode (`/sdlc:plan`).** ## Do not - Do not write code or an implementation plan (files, order of work). That is Stage 3. - Do not silently resolve a policy conflict. Flag it. - Do not mark `status: accepted`.
From bibutikoley/ai-sdlc-claude, MIT.
More from bibutikoley/ai-sdlc-claude
- 0.564score
review
bibutikoley
Stage 5 (Deploy) AI review of the AI-native SDLC.
SkillBFreshness: B★ 3Claude Code +9 - 0.564score
setup
bibutikoley
Install the AI-native SDLC loop into a product repository.
SkillBFreshness: B★ 3Claude Code +9 - 0.564score
eval-add
bibutikoley
Stage 4 (Test) continuous evals.
SkillBFreshness: B★ 3Claude Code +9 - 0.564score
incident
bibutikoley
Stage 6 (Maintain) of the AI-native SDLC.
SkillBFreshness: B★ 3Claude Code +9 - 0.564score
intent
bibutikoley
Stage 1 (Plan) of the AI-native SDLC.
SkillBFreshness: B★ 3Claude Code +9 - 0.564score
metrics
bibutikoley
Compute and explain the AI-native SDLC's leading and lagging indicators from Git history, PR metadata and CI (intent time-to-commit, intent→spec, spec→plan…
SkillBFreshness: B★ 3Claude Code +9 - 0.564score
plan
bibutikoley
Stage 3 (Build) entry of the AI-native SDLC.
SkillBFreshness: B★ 3Claude Code +9 - 0.564score
brand-guidelines
bibutikoley
The organisation's brand and content standards applied to specs and UI work.
SkillBFreshness: B★ 3Claude Code +9 - 0.564score
compliance-policy
bibutikoley
Regulatory and change-management obligations as constraints on specs, plans, releases and reviews.
SkillBFreshness: B★ 3Claude Code +9 - 0.564score
tracker-sync
bibutikoley
Keep the legacy tracker (Jira, YouTrack, GitLab issues, GitHub issues) and the Markdown artifacts linked according to sdlc.config.json sourceOfTruth (repo |…
SkillBFreshness: B★ 3Claude Code +9 - 0.564score
security-policy
bibutikoley
The organisation's security policy as constraints on specs, plans, code and reviews.
SkillBFreshness: B★ 3Claude Code +9 - 0.564score
ux-standards
bibutikoley
Interaction, accessibility and content-structure standards for anything users touch.
SkillBFreshness: B★ 3Claude Code +9
Signals
- Stars
- 3
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Sep 5, 2026 (1 month ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/bibutikoley/ai-sdlc-claude
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from bibutikoley
- 0.564score
plan
bibutikoley
Stage 3 (Build) entry of the AI-native SDLC.
SkillBFreshness: B★ 3Claude Code +9 - 0.564score
eval-add
bibutikoley
Stage 4 (Test) continuous evals.
SkillBFreshness: B★ 3Claude Code +9 - 0.564score
incident
bibutikoley
Stage 6 (Maintain) of the AI-native SDLC.
SkillBFreshness: B★ 3Claude Code +9 - 0.564score
intent
bibutikoley
Stage 1 (Plan) of the AI-native SDLC.
SkillBFreshness: B★ 3Claude Code +9 - 0.564score
metrics
bibutikoley
Compute and explain the AI-native SDLC's leading and lagging indicators from Git history, PR metadata and CI (intent time-to-commit, intent→spec, spec→plan…
SkillBFreshness: B★ 3Claude Code +9 - 0.564score
brand-guidelines
bibutikoley
The organisation's brand and content standards applied to specs and UI work.
SkillBFreshness: B★ 3Claude Code +9
Cite this page
spec skill. Agent Mart. https://agentmart.dev/skills/bibutikoley-spec (last updated 2026-10-24)