AgentStack
Browse Sign in
Browse Why AgentStack Sell Docs
Sign in
SKILL verified MIT Self-run

Spec Dev

skill-vaayne-agent-kit-spec-dev · by vaayne

>

— No reviews yet
0 installs
44 views
0.0% view→install

Install

$ agentstack add skill-vaayne-agent-kit-spec-dev

✓ scanned · ✓ verified, works with Claude Code, Cursor, and more.

Security review

✓ Passed

No 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.

View the full security report →

Verified badge

Passed review? Show it. Paste this badge into your README, it links to the public security report.

AgentStack Verified badge Links to your public security report.
[![AgentStack Verified](https://agentstack.voostack.com/badges/verified.svg)](https://agentstack.voostack.com/security/report/skill-vaayne-agent-kit-spec-dev)

Reliability & compatibility

✓ Security review passed
0 installs to date
— no reviews yet
● 3mo ago

Declared compatibility

Claude CodeClaude Desktop

Compatibility is declared by the source manifest. End-to-end runtime verification is coming, see below.

Preview Execution monitoring

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 →
Are you the author of Spec Dev? Claim this listing to set pricing, connect Stripe payouts, and keep 70% of every sale.
Sign up to claim

About

The flow is Grill → Plan → Impl (phase by phase). The only artifact is plan.md, which lives at ~/.agents/sessions/{project}/{date}-{feature}/plan.md where {project} is the git repo name (if in a git repo) or the basename of the current working directory, {date} is YYYY-MM-DD, and {feature} is a short kebab-case label. That one file holds the plan, inline reviews, impl questions, and per-phase handoff notes.

Grill

Stress-test the design before writing anything down. Read [grilling-guide.md](./references/grilling-guide.md) — interrogate one question at a time, sharpen fuzzy terms, cross-reference with code, probe reversibility.

Skip grilling when the approach is already clear — small bug fixes, straightforward changes, or work where the plan is obvious don't need interrogation.

Plan

Write plan.md following [plan-template.md](./references/plan-template.md). The key principle: explain why, not just what. For each significant decision, cover what you decided, what alternatives you ruled out, and what tradeoffs you accepted. Break the work into phases, and give every phase an Acceptance block defining "done".

Inline reviews use the > quote + **Review (name):** pattern; resolve with **Resolved:** and update the plan text above. See [review-patterns.md](./references/review-patterns.md).

Impl (phase by phase)

Each phase is its own catch-up → execute → verify → review-if-needed → commit → handoff cycle. Full workflow in [implementation-guide.md](./references/implementation-guide.md).

Two things to know up front:

  • Review gates — pause for user approval after grilling (before writing the plan) and after the plan (before implementation). Within impl, don't gate every phase, but escalate on surprises or major rework.
  • Per-phase review is a judgment call — for each phase, decide whether the work needs another pass before moving on. Trivial phases self-review; phases touching critical paths, security, data, or non-obvious logic warrant a closer look. When in doubt, review.

Impl questions raised mid-phase use the **Question (name):** / **Answer:** pattern from [review-patterns.md](./references/review-patterns.md). When an answer changes the plan, update the affected section.

Quick reference

| Action | What to do | Reference | | ------------------ | -------------------------------------------------------------------------------- | --------------------------------------------------------------- | | Grill a design | Interrogate one question at a time, sharpen terms, cross-reference code | [grilling-guide.md](./references/grilling-guide.md) | | Write plan | Create ~/.agents/sessions/{project}/{date}-{feature}/plan.md | [plan-template.md](./references/plan-template.md) | | Add review | > quote then **Review (name):** | [review-patterns.md](./references/review-patterns.md) | | Resolve review | **Resolved:** after the review block; update plan text above | [review-patterns.md](./references/review-patterns.md) | | Ask impl question | **Question (name):** indented under the blocked task | [review-patterns.md](./references/review-patterns.md) | | Answer question | **Answer:** directly below the question | [review-patterns.md](./references/review-patterns.md) | | Run a phase | Catch up on handoff → implement → verify Acceptance → review? → commit → handoff | [implementation-guide.md](./references/implementation-guide.md) | | Check open threads | grep -n "Review\|Resolved" plan.md and grep -n "Question\|Answer" plan.md | |

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

Install and usage instructions live in the source repository linked above.

Reviews

No reviews yet, be the first.

Versions

  • v0.1.0 Imported from the upstream source.