Agent skill

full_auth

When explicitly authorized, execute end-to-end with a phased plan in one run until Definition of Done is met.

Stars 163
Forks 31

Install this agent skill to your Project

npx add-skill https://github.com/majiayu000/claude-skill-registry/tree/main/skills/devops/full-auth

Metadata

Additional technical details for this skill

short description
Full authorization execution mode

SKILL.md

Authorization

Use this skill only when the user explicitly grants permission to edit files and complete the task end-to-end without asking “should I continue?”.

Execution Mode

PHASED PLAN, ALL PHASES IN ONE RUN

  • Produce a phased implementation plan first (Phase 1/2/3…).
  • Immediately execute all phases in sequence in the same run, until Definition of Done is met.
  • Only pause if blocked by missing info or if a command fails and cannot be resolved autonomously.

Blocking Rule

If blocked, ask exactly one concise question that would unblock the work. Otherwise, keep going.

Definition of Done (generic)

  • Safety validation passes (respect .agentignore).
  • Use the best available repo verification:
    • Prefer .agent-docs/memory/COMMANDS.json and MANIFEST.yaml.
    • If missing, run capability/command detection and record confidence.
    • Run tests, lint, and build as applicable.
  • If QUALITY_GATES.md exists, follow it as the source of truth.
  • Provide a final summary and list all modified/created files.

Output Requirements

  1. Investigation notes (what matters in the repo).
  2. Phased plan (Phase 1/2/3… with files and behaviors).
  3. Implementation (commit-style narrative by phase).
  4. Files changed + verification results (commands run and outcomes).

Guardrails

  • Always respect .agentignore forbidden zones.
  • Enforce Trust Layer TDD for behavior changes.
  • Keep an Action Ledger entry for meaningful work.

Didn't find tool you were looking for?

Be as detailed as possible for better results