Install
$ agentstack add skill-enocgit-sdlc-kit-feature-start ✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
Verified badge
Passed review? Show it. Paste this badge into your README, it links to the public security report.
Reliability & compatibility
Declared compatibility
Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.
We're building live execution health for every listing: tool-call success rate, median latency, uptime, and last-checked timestamps, measured, not self-reported. It isn't live yet, so we don't show numbers we can't stand behind.
How agent discovery & health will work →About
feature-start
Prepare a clean, well-contextualized starting point for implementing one task, then stop for a plan approval gate.
Steps
- Pick the task. Confirm the GitHub issue # and slug with the user. One task per run.
- Isolate the workspace — default to a feature branch. Create
feat/{issue#}-{slug}directly.
Only reach for an isolated git worktree (delegating to using-git-worktrees if installed) when isolation is genuinely critical — parallel work on multiple features, or disposable experiments — or when the user asks. Either way, FIRST run these guards (the worktree skill does them for you; the branch path must not skip them):
- Clean tree:
git status --porcelainis empty — nothing uncommitted to clobber. - Fresh base: check out the default branch and pull, so the feature branches off latest.
- Green baseline: install deps and run the test suite once; if it's already red, stop and
report — don't start work on a broken baseline. Then create feat/{issue#}-{slug}. One feature per workspace; don't touch other branches.
- Load context. Read into context:
docs/context.md(domain, glossary, hard constraints — incl. retro learnings from prior cycles)- The feature's PRD in
docs/prd/ - Any ADR(s) it depends on in
docs/adr/ - The frozen contract artifact (OpenAPI/tRPC/schema) the task implements against
docs/test-strategy.md(Definition of Done + which layer to test at)
- Plan. Enter plan mode. Produce a granular plan (small steps, exact file paths, the tests
you'll write first). Hand off to executing-plans / test-driven-development for execution.
- GATE. Present the plan. Ask for approval before writing any code.
Rules
- Implement strictly against the frozen contract. If you discover the contract is wrong, STOP
and raise it — a contract change is a decision (new/updated ADR), not an inline edit.
- If the task is bigger than ~a day of work, propose splitting it before starting.
- Keep the SDLC status header on every user-facing message, just like the conductor — open with
SDLC ▸ Stage 4/8 Implement · task #N · {next: plan approval / …}. You're inside the pipeline even though sdlc isn't the active skill; don't drop the header once implementation starts.
- Don't commit or push unless asked.
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: enocgit
- Source: enocgit/sdlc-kit
- License: MIT
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet, be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.