Scope guard
Activates at the start of a Claude Code session when the user's first message describes a task or coding goal. Triggers on opening messages containing: fix, build, implement, refactor, add, update, create, debug, test, review, deploy, migrate, optimize, write, remove, rename. Also triggers on: "set scope", "define contract", "what's in scope", "let's work on", "I need to". Does NOT re-trigger mid-session. Does NOT trigger on questions or follow-ups in an ongoing session. Only fires on the opening task-description message.From its SKILL.md
npx -y skills add aarshsharma2803-stack/optimusprime --skill scope-guardAssembled from the repository path, not quoted from the project. Check it against their README if it does not work.
One thing to look at
- 0 stars0 stars. Stars are a popularity signal and not a quality one, but at this level it is likely that nobody has read this closely except its author, and you would be relying on your own review.
SKILL.md
1.3 KB, 213 tokens by cl100k_base, as published. Nobody here has run it
Silently extract scope contract from opening message. Walk up from cwd to find .optimusprime/. If not found, skip.
Write .optimusprime/contract.json:
goal: verb-first one sentence ("Implement X", "Fix Y in Z")in_scope: file globs inferred from taskout_of_scope: always include["node_modules/**", ".env", ".git/**", "__pycache__/**"]; add others if obviouscomplexity_budget:"minimal"(1–2 files) /"moderate"(3–10 files) /"full"(cross-cutting)agent_id:"main"created_at: ISO 8601 UTCsession_id: random 8-char hex
If goal is clear: write contract, output nothing, proceed to task. If genuinely ambiguous: ask exactly ONE question to resolve scope, then write.
After writing: no output. Proceed directly to the user's task.
What ships with it
Read from the repository
Just SKILL.md. No reference files, no scripts.