Menu
SkillsMP has collected 10 skills from yezhwi/superpowers-engineering-harness. Open a skill to review its source and details.
Showing 10 of 10 collected skills.
Use when starting or resuming a task in a repo with .harness/ state, when unsure which harness phase comes next, or when asked to run the Engineering Harness loop end-to-end. Orchestrates the fixed state machine; MUST NOT implement business code itself.
Review Q2/Q3 changed business paths for production diagnosability and persist structured review input.
Turn a User Requirement into a Task Contract: Acceptance Criteria, Requirements, Invariants, Risks, Verification Plan. Produces .harness/requirements.yaml and .harness/invariants.yaml and advances task state CREATED -> SPECIFYING -> PLANNED. This skill MUST…
Adversarial review of a Task Contract: construct failure scenarios, attempt to violate Requirements and Invariants. Produces .harness/findings/FND-nnn.yaml files with status=PROPOSED only. This skill MUST NOT confirm bugs, review code style, or suggest…
Reproduce an adversarial finding through the mandatory lifecycle PROPOSED -> REPRODUCING -> CONFIRMED (bug reproduced, test RED) -> FIXING -> FIXED (test GREEN) -> VERIFIED (full regression) -> CLOSED, or REJECTED. Each status has one meaning and its own…
Review verified diff for unnecessary implementation complexity before adversarial review.
Run before implementation to record minimal necessary approach using Harness Decision Ladder.
Use when quality_gate.py has returned PASS or BLOCKED and the loop must decide to finish (CONVERGED/DONE), continue iterating, or escalate. Applies v0.1 escalation rules with no scoring.
Use when entering VERIFYING or before GATING and fresh evidence is required. Decides which evidence types the task needs, then runs collect_evidence.py; MUST NOT judge pass/fail itself.
Use when task state is GATING and a go/no-go decision is needed before CONVERGED. Runs quality_gate.py and maps its exit code to state transitions; MUST NOT assess quality by judgement.