Loading...
Loading...
Found 16 Skills
Shiplight QA toolkit — the single entry point for all Shiplight test/QA work. Use ONLY when the user explicitly says 'shiplight' (e.g. 'write a shiplight test', 'use shiplight to verify X', 'shiplight cover') or invokes /shiplight. Routes to subcommands: init, auth, update, create-yaml-tests, create-agent-verification, cover, fix, verify, review, ci, cloud, support, help.
Cut an OpenSEO release — bump the version, draft user-facing release notes from commits since the last tag, run a review + subagent-verification pass, and open a "release: vX.X.X" PR. Use when the user asks to prepare a release, bump the version, or write release notes.
Use when an implementation task has just been completed and the user wants independent verification that the work is actually correct. Dispatches two subagents that perform the SAME verification checks independently, then merges their findings into one deduplicated report with severity and confidence. Triggers when the user says "verify this", "double-check the implementation", "cross-check what was done", "did this actually work", "verify the task", or invokes a slash command for verification. Especially valuable right before commits, PRs, or hand-offs, where over-claiming is most costly.
Execute a task with sub-agent implementation and LLM-as-a-judge verification with automatic retry loop
Multi-agent adversarial verification with convergence loop. Two independent review agents must both pass before output ships.
Comprehensive verification with parallel test agents. Use when verifying implementations or validating changes.
Create a structured format for documenting feature requirements as user stories. JSON files with testable acceptance criteria that AI agents can verify and track.
Integrate oh-my-ag with MCP for ulw-style multi-agent workflows. Covers install, setup, bridge mode, and verification steps.
Verifies the agent's current work against a specific question by analyzing unstaged changes, staged changes, recent commits, and codebase context. Answers succinctly for a senior audience. Use when user says "/check", "verify that", "confirm that", "check if", "is X done?", or asks about current session changes.
Runs a doer -> verifier-panel -> consensus loop to verify a deliverable before it ships. An orchestrator freezes acceptance criteria before implementation, dispatches a doer, then convenes a context-walled panel of independent verifiers - including an adversary with an explicit must-oppose mandate - for evidence-anchored review adjudicated to a SHIP / SHIP_WITH_CAVEATS / ITERATE / BLOCK / ESCALATE verdict logged to a ledger. Use for multi-agent verification of any artifact - code slices, plans, documents, audits - whenever asked to verify a deliverable, vet a plan, run a consensus review or independent review, set up a doer-verifier loop, or gate a ship decision. Works on any platform with parallel subagents; degrades to sequential fresh-context sessions without them. Not for trivial single-file edits or ordinary code review.
LLM-as-judge evaluation framework with 5-dimension rubric (accuracy, groundedness, coherence, completeness, helpfulness) for scoring AI-generated content quality with weighted composite scores and evidence citations
Epistemic verification framework for AI-generated assertions. Requires evidence before acting on LLM claims about code behavior, system state, API responses, or factual statements. Use when an AI agent makes claims that will drive decisions, before acting on research results, or when an agent asserts something is true without showing evidence.