workflow-tournament-053843 — independently scanned and version-tracked by SaferSkills.
SaferSkills independently audited workflow-tournament-053843 (Agent Skill) and scored it 100/100 (green). The audit ran 55 deterministic rules across Security, Supply Chain, Maintenance, Transparency, and Community; it found 0 high-severity and 0 lower-severity findings. The full rule-by-rule trace and per-finding evidence are below. Free, methodology-open.
Findings & checks · 0 flagged
Every scanned point with the score it earned and what moved between them.
First recorded scan — no prior version to compare against.
The primary manifest — the file an agent reads to learn what this artifact does.
Execute a tournament-style parallel implementation: $ARGUMENTS
You are executing a parallelized "tournament" implementation workflow. Instead of decomposing a feature into smaller tasks for individual agents, you will create multiple isolated Git worktrees. You will dispatch one sub-agent to each worktree to build the entire feature independently. Once all sub-agents finish, you will review their implementations, select the best one along with the user, and merge it back into the primary feature branch.
This approach maximizes exploration of different implementation strategies and keeps the canonical codebase cleanly isolated until a winner is chosen.
Before starting, verify:
feat/X or fix/Y).git status).For each of the N sub-agents (e.g., from 1 to N):
git branch <current-branch>-agent-<N>.worktrees/ directory (ensure .worktrees is in .gitignore): mkdir -p .worktrees
git worktree add .worktrees/agent-<N> <current-branch>-agent-<N> cd .worktrees/agent-<N>
pnpm install
# Copy any necessary environment files (e.g., cp ../../.env ./)Dispatch a sub-agent to each initialized worktree concurrently.
For each sub-agent:
.worktrees/agent-<N>).Wait for all sub-agents to complete their implementations. If a sub-agent fails or times out, note the failure but do not halt the overall process.
Collect the summaries, the final git log, and git diff from each worktree. Ensure each sub-agent commits their work to their respective <current-branch>-agent-<N> branch.
Before evaluating the elegance or architecture of any implementation, apply a strict first-pass deterministic filter. This should filter out ~80% of the unviable options automatically.
For each sub-agent's worktree, determine if they pass the following binary gates:
pnpm test (or the project's test command) pass entirely?pnpm run build)?pnpm run typecheck or equivalent)?Disqualify any sub-agent that fails these deterministic gates. Only implementations that pass the objective binary gates are allowed to move on to the final subjective review.
Review the implementations from the remaining successful sub-agents. Evaluate them based on:
CRITICAL RULE: The human is the final decider by default. While you must present your analysis and recommend a winner based on the criteria above, you MUST NOT merge any branch until the human explicitly reviews the options and approves a winner.
git merge <current-branch>-agent-<WINNER_N> git worktree remove .worktrees/agent-* --force
git branch -D <current-branch>-agent-1 ...Present your final summary to the user:
## Tournament Implementation Complete
### Selected Winner: Agent <N>
- **Reasoning**: [Why this implementation was chosen over the others]
### Participants
- Agent 1: [Status/Brief Summary]
- Agent 2: [Status/Brief Summary]
...
### Next Steps
- The winning code has been merged into `<current-branch>`.
- Worktrees have been cleaned up.
- Ready for final review and push.~30 seconds. Free. No account. Every finding cites a rule and a line of evidence.