brutal-spec
Synthesize repo evidence and resolved requirements into a reviewed, decision-complete implementation specification.
Skill by sssemil
Chalk Labs
Need help deploying brutal-spec?
Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.
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/sssemil-brutal-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
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/brutal-spec
What's inside
- Skill name
brutal-spec- Description
- Synthesize repo evidence and resolved requirements into a reviewed, decision-complete implementation specification. Use after requirements grilling, when converting a planning conversation or completed investigation map into the parent specification for a Brutal plan.
- License
- MIT
- Instructions
- 67 lines in SKILL.md · view on GitHub
Files (2)
- SKILL.md
- openai.yaml · agents/
SKILL.md (first lines)
# Brutal Spec Turn a decision-complete requirements brief into the behavioral and technical contract that implementation tickets will inherit. Do not publish artifacts. ## Verify The Inputs Read the requirements brief, relevant repository instructions, domain glossary, ADRs, affected interfaces, tests, and any linked research, prototype, or completed investigation map. Return unresolved material decisions to `brutal-grill`; do not hide them as assumptions. ## Draft The Specification Write only the contract an implementer needs. Reuse the requirements brief by reference instead of restating it, combine related material, and omit sections that do not change implementation. Use this compact shape: ```markdown ## Outcome And Scope ## Behavior And Boundaries ## Interfaces, Data, And Migration ## Implementation Decisions ## Verification And Acceptance ## Risks, Assumptions, And Domain Documentation ``` State public interface, schema, migration, and wire-format changes precisely. Prefer stable module and domain names over file paths. Include a code or type snippet only when a prototype established a decision more precisely than prose. Choose the highest existing public seams that can verify the behavior. Keep the number of seams small, explain any new seam, and name exact repository commands that will verify the finished work. Propose a glossary edit when planning resolves domain language. Propose an ADR only when the decision is hard to reverse, surprising without context, and the result of a genuine tradeoff. Do not edit either document yet. ## Stress-Test The Draft When independent reviewers are available, review the same draft and evidence from these perspectives: - architecture, module depth, and coupling - data model, compatibility, and migration safety - public interface, UX, and integration behavior - reliability, security, operations, and delivery risk Classify findings as `PLAN BLOCKER`, `IMPLEMENTATION NOTE`, or `SUGGESTION`. Validate each finding against the repository, fold valid notes into the draft, and return every simultaneously answerable validated blocker to `brutal-grill` in one round. Defer a blocker when another answer can change its necessity, wording, or options; never expose an unvalidated finding as a user question. ## Approval Gate Present outcome, scope, key decisions, material risks, and acceptance criteria as the compact approval view. Keep that view to one short paragraph and at most seven bullets total; group related decisions, risks, and criteria. Show the complete revised specification only when the user asks for it or the specification text itself is the requested deliverable. The stage completes only when the user approves and no `PLAN BLOCKER` or open material decision remains. Return the approved full specification to the calling skill without persisting it.
From sssemil/SKILLS, MIT.
More from sssemil/SKILLS
- 0.512score
design-jepa-models
sssemil
Design, review, prototype, and evaluate JEPA models, joint-embedding predictive architectures, V-JEPA/I-JEPA-style representation learners, action-conditioned…
SkillCFreshness: C★ 2Claude Code +9 - 0.509score
brutal-pr-review
sssemil
Review an open pull request with Brutal multi-perspective scrutiny, snapshot-safe BRUTAL.md context, and provider-native findings.
SkillCFreshness: C★ 2Claude Code +9 - 0.509score
brutal-swarm
sssemil
Drain a BRUTAL.md task graph through parallel isolated workers and stacked pull requests.
SkillCFreshness: C★ 2Claude Code +9 - 0.507score
brutal-pr-fix-loop
sssemil
Run a bounded provider-neutral pull-request review and finding-fix loop for up to three passes.
SkillCFreshness: C★ 2Claude Code +9 - 0.507score
funcvec-dedupe-fixer
sssemil
Use when the user wants Rust duplicate-function detection and cleanup with funcvec, especially crate-by-crate scans, duplicate triage, boilerplate refactors…
SkillCFreshness: C★ 2Claude Code +9 - 0.507score
brutal-grill
sssemil
Repo-grounded requirements interrogation that resolves related planning decisions in batched rounds.
SkillCFreshness: C★ 2Claude Code +9 - 0.507score
brutal-plan
sssemil
Orchestrate repo-grounded grilling, specification review, cohesive ticket design, and work-store publication.
SkillCFreshness: C★ 2Claude Code +9 - 0.507score
brutal-reality
sssemil
Stress-test business ideas against the "Reality's Moat" scar tissue framework.
SkillCFreshness: C★ 2Claude Code +9 - 0.507score
brutal-tickets
sssemil
Decompose an approved specification into the smallest useful dependency graph of cohesive, decision-complete implementation scopes.
SkillCFreshness: C★ 2Claude Code +9 - 0.507score
codex-coop
sssemil
Co-operative workflow with OpenAI Codex CLI for plan review and pre-commit code review.
SkillCFreshness: C★ 2Claude Code +9 - 0.504score
brutal-pr-finding-fixer
sssemil
Fix generated pull-request finding occurrences through the BRUTAL.md code-host adapter, one focused change at a time, with complete queue scans for…
SkillCFreshness: C★ 2Claude Code +9 - 0.504score
brutal-inf-fix-loop
sssemil
Review and fix an open pull request without a pass limit until a fresh review has no CRITICAL or MAJOR findings.
SkillCFreshness: C★ 2Claude Code +9
Signals
- Stars
- 2
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- Jul 22, 2026 (3 months ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/sssemil/SKILLS
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from sssemil
- 0.512score
design-jepa-models
sssemil
Design, review, prototype, and evaluate JEPA models, joint-embedding predictive architectures, V-JEPA/I-JEPA-style representation learners, action-conditioned…
SkillCFreshness: C★ 2Claude Code +9 - 0.509score
brutal-swarm
sssemil
Drain a BRUTAL.md task graph through parallel isolated workers and stacked pull requests.
SkillCFreshness: C★ 2Claude Code +9 - 0.509score
brutal-pr-review
sssemil
Review an open pull request with Brutal multi-perspective scrutiny, snapshot-safe BRUTAL.md context, and provider-native findings.
SkillCFreshness: C★ 2Claude Code +9 - 0.507score
brutal-reality
sssemil
Stress-test business ideas against the "Reality's Moat" scar tissue framework.
SkillCFreshness: C★ 2Claude Code +9 - 0.507score
brutal-pr-fix-loop
sssemil
Run a bounded provider-neutral pull-request review and finding-fix loop for up to three passes.
SkillCFreshness: C★ 2Claude Code +9 - 0.507score
brutal-plan
sssemil
Orchestrate repo-grounded grilling, specification review, cohesive ticket design, and work-store publication.
SkillCFreshness: C★ 2Claude Code +9
Done for you
Rather not set it up yourself?
Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.
See how we build AI agents →Read the case studies
Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.
- launches shipped
- 50+
- launches shipped
- media placements
- 500+
- media placements
- founders on every account
- 2
- founders on every account
Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block
Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.
Get brutal-spec running in your business
Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.
Cite this page
brutal-spec skill. Agent Mart. https://agentmart.dev/skills/sssemil-brutal-spec (last updated 2026-10-24)