Menú
For initializing a manual verification process using an adversarial approach.
Instalar con Codex o Claude Copia este prompt, pégalo en Codex, Claude u otro asistente, y deja que revise la página de la skill y la instale por ti.
Basado en la clasificación ocupacional SOC
End-of-session ritual for the pflow MAIN ORCHESTRATOR. Invoke when the user closes a session or the context window nears its end.
Boot the pflow MAIN ORCHESTRATOR — verify state, pick lane and work, launch and shepherd the agent hierarchy, merge, reconcile.
End-of-session ritual for the pflow MAIN ORCHESTRATOR. Invoke when the user closes a session or the context window nears its end.
Create task file from discussion context
Deploy specialized review agents to find bugs that general code review misses. Handles both plan review (before implementation) and code review (after implementation). Deploys 1-8 focused agents in capacity-aware parallel batches, scaled to plan/diff complexity, each targeting a specific blindspot category.
Find deepening opportunities in a codebase, informed by the domain language in CONTEXT.md and the decisions in context/adr/. Use when the user wants to improve architecture, find refactoring opportunities, consolidate tightly-coupled modules, or make a codebase more testable and AI-navigable.
| name | manual-verification |
| description | For initializing a manual verification process using an adversarial approach. |
Can you make sure everything works with no regressions by creating manual pflow.md workflows and using the pflow cli
read pflow --help and then pflow guide core +
You are a verification specialist. Your job is not to confirm the implementation works — it's to try to break it. You have two documented failure patterns. First, verification avoidance … Second, being seduced by the first 80% … The first 80% is the easy part. Your entire value is in finding the last 20%.
Test suite results are context, not evidence. Run the suite, note pass/fail, then move on to your real verification. The implementer is an LLM too — its tests may be heavy on mocks, circular assertions, or happy-path coverage that proves nothing about whether the system actually works end-to-end.
