one-person-lab

Design domain-neutral Foundry EvalSpec cases, gates, protected-test requirement categories, failure taxonomies, and EvidenceBundle expectations. Use when creating or reviewing evaluation semantics for an AgentBlueprint without taking over protected tests, execution, qualification, or Owner acceptance.

Interpret OPL FoundryRun, EvidenceBundle, qualification, risk, Owner-gate, canary, activation, and rollback evidence. Use when an operator needs an evidence-bound failure classification or next route for an Agent create, takeover, or improve run.

Design OPL Stagecraft support for stage goal, prompt shaping, rubric, capability-use decisions, route-back reasoning, handoff, and receipt/blocker lower-bound design. Use when authoring or reviewing an OPL stage strategy or stage prompt wrapper while keeping Stagecraft contracts limited to descriptors, allowed/forbidden surfaces, and no-authority receipt/blocker lower bounds.

Use when reviewing OPL family stage assumption lifecycle projections, stale or missing assumptions, monitor gaps, owner gaps, minimal counterexamples, repair actions, and no-authority route-back recommendations.

Use when reviewing OPL stage candidate portfolio refs-only projections, candidate ref coverage, assumption/provenance/negative-path/advisory-metric refs, human-review route-back, missing refs, and domain-authority overclaim risks.

Use when reviewing OPL family stage-pack source specs and replay certifications, including body-free source diffs, visual-equivalent spec gaps, replay blockers, missing runtime event refs, missing receipt refs, and owner-route workorders.

Use when critiquing OPL Stagecraft quality gates, stage admission projections, stage output evidence, evaluator framing, rubric gaps, trust lanes, composition obligations, failure localization, human-review burden, owner-route coverage, route-back packets, and no-authority gate recommendations.

Use when writing or reviewing OPL Workspace handoff packets, source-readiness audit sections, source refs, artifact refs, workspace briefings, missing-input route-back notes, and owner-route packets while keeping locator, path, artifact-unit, and source-readiness truth in Workspace or program-owned surfaces.

Use for OPL Connect external skill/source search-inspect-sync routing, connector receipt debugging, single-skill sync decisions, and source refs/no-authority review. Helps Codex decide whether an external source or skill is needed, which candidate to inspect or sync, whether connector metadata or receipts are too risky, and how to hand candidate refs to a domain owner without claiming domain truth, quality verdict, owner acceptance, or readiness.

Use when the default OPL/domain professional skills do not cover a specialist tool, source, workflow, or method and Codex should route through OPL Connect registered external-skills search/inspect before optional single-skill workspace/quest sync. Keeps large external skill libraries out of default context and returns candidate refs, support map, and route recommendation without owner receipts, typed blockers, domain verdicts, or readiness claims.

Use when reviewing OPL agent package trust plus lifecycle: manifest digest, provenance, dependency refs, carrier exposure, install/update/repair/rollback, dirty/developer/manual-required states, Codex reload proof, and owner routing.

Use when reviewing OPL Charter or authority-boundary materials, including policy, ADR, RFC, closeout reports, owner split, no-second-truth risks, forbidden claims, readiness overclaim, authority language, and completion or ready claims before publication or handoff. This skill prepares AI authority findings only; it does not change policy source of truth or sign owner answers.

Use when interpreting OPL Console/App operator projections, current_owner_delta, action catalogs, first-run readiness wording, Settings IA, Runtime page task awareness, user workbench actions, release evidence refs, and forbidden claims without executing actions or claiming App, domain, runtime, release, or readiness status.

Use when triaging OPL stalls, currentness drift, runtime/readback mismatch, heartbeat alerts, provider failures, owner-route gaps, repeated blocker/nonprogress loops, conflict/blocker envelopes, duplicate tasks, human gates, retry/dead-letter states, or false progress/readiness incidents. Applies L0-L4 root-cause depth and routes repair without writing authority surfaces.

Use when curating OPL memory/artifact/local-data lifecycle refs, artifact-unit handoff, provenance, cleanup, retention, storage archive/restore evidence, log/cache cleanup, and owner-route briefs without writing artifact, memory, or local-data authority.

Use when reviewing OPL pack admission candidates, declarative pack boundaries, capability/authority ABI, tool affordance, generated or hosted surface inputs, registry fit, contract evidence, owner routes, allowed writes, forbidden authority, and route-back decisions before a pack is admitted.

Use when auditing OPL runtime/provider environment fit, runtime environment bundles, native helper diagnostics, long-soak evidence, recovery attempts, provider observations, no-regression refs, and route-back decisions without mutating runtime or provider state.

Use when reviewing OPL source-module boundaries, public entrypoints, dependency direction, forbidden imports, owner alignment, source placement, shell/upstream capability intake, and domain private-tail retirement routes without changing source truth.

Use when routing across OPL agent, capability, tool-card, source, and owner catalogs; choosing owner, source, skill, or connector refs; diagnosing catalog ambiguity; and preparing refs-only route packets. Helps keep catalog registry, lifecycle index, and refs graph as program authority.

Use when curating refs-only evidence, receipts, provenance, closeout proof, claim support, and evidence gap classifications for OPL work while preparing evidence briefs without writing append-only ledgers or signing receipts.

Use when reviewing OPL owner evidence intake packets, observed refs, acceptance candidates, owner-chain scaleout evidence, and evidence class fit without signing owner acceptance.

Use when reviewing opl quality details --json, Sentrux output, or line-budget sidecars to decide whether code quality remediation is needed, how to make the smallest fix, and when findings stay advisory.

Use when auditing OPL Connect connector receipt candidates, normalized source refs, invocation evidence, failed-provider records, registry fit, owner routes, and no-authority handoff briefs.

Use when operating or debugging OPL Runway compute support, including environment setup diagnosis, SSH or SLURM access, Modal or managed endpoint selection, provider receipt review, Runway handoff briefing, and provider failure classification. Helps choose a compute/provider route and prepare operator handoff/debug notes without claiming provider, live runtime, endpoint, queue, or readiness authority.

Use when writing or reviewing OPL Plan Completion Audit reports for full-plan, all-the-way, closeout, or completion claims. Helps map plan items to done/partial/not_started/blocked status, evidence class, gaps, next owner, and forbidden readiness or authority claims.