Total Skills
18
Skills published by sformisano with real stars/downloads and source-aware metadata.
Total Skills
18
Total Stars
0
Total Downloads
0
Comparison chart based on real stars and downloads signals from source data.
change-walkthrough
0
changelog-authoring
0
execution-discipline
0
lifecycle-coherence-audit
0
rust-code-review
0
squad-implementation-verification
0
squad-implementer
0
squad-lead
0
Use when explaining a code changeset in chat, MR descriptions, or explicitly requested local delivery notes. Orders concepts before files and enforces full code-diff coverage with surface-specific reference formats.
Use only when the user explicitly asks for a local changelog or delivery note, or when reading existing local scratch notes. This skill is not a delivery, review, MR, or commit gate.
Foundational execution contract for agent work. Covers constraint precedence, lifecycle integrity, preflight checks, and external tool failure handling.
Audit governance artifacts for a formal squad lifecycle before resume, delivery-ready, or delivered claims. Use when manifest, handoff, status, or resume docs disagree inside a squad journal.
Use for external Rust review commands such as local delta, branch diff, merge request, or codebase audit. Prioritizes correctness, regression risk, async safety, compatibility, and test depth.
Defines the squad lifecycle implementation report contract — claim proof results, fixture use, exercise setup, scenario coverage, regression baseline, parity evidence, and deferred items. Use when writing or auditing a squad implementation artifact while acting as squad implementer, review author, or lead, not for ordinary code review or status writeups.
Specialist for executing an approved squad plan phase or squad review findings and writing an implementation report. Use when dispatched by squad-lead with plan, triage, phase, and artifact paths, not for ordinary coding.
Coordinates formal squad lifecycle runs with requirement, triage, manifest, plan, implementation, review, adversary gates, and delivery handoff. Use when the user explicitly wants squad/protocol orchestration, mission-critical multi-phase delivery that needs formal squad gates, or resume of a squad journal, not for ordinary coding or generic task planning.
Defines the manifest.yaml state file for formal squad lifecycle runs. Use when acting as squad lead to create, resume, validate, or update structured task state, not for ordinary project manifests.
Specialist for authoring squad lifecycle plan artifacts from a lead-provided requirement, triage, manifest, and output path. Use when dispatched by squad-lead or when explicitly asked to write a squad plan artifact.
Specialist for critiquing a squad plan artifact for scope, verification, phase, and boundary defects. Use when dispatched by squad-lead or explicitly asked to review a squad plan artifact.
Defines the required shape of a squad lifecycle plan artifact. Use when loaded by squad plan author, squad plan critic, or squad lead to write or check a squad plan, not for ordinary coding plans.
Specialist for adversarially stress-testing a converged squad review after review-critic returns SATISFIED with action submit. Use when dispatched by squad-lead at Phase 3.5 or 4.5, not for ordinary review.
Specialist for producing a squad review artifact for a completed implementation round. Use when dispatched by squad-lead with implementation, plan, triage, diff, and output paths, not for casual code review.
Specialist for auditing a squad review artifact for missed findings, false positives, weak evidence, and routing mistakes. Use when dispatched by squad-lead or explicitly asked to critic-check a squad review.
Defines the required structure of a squad lifecycle review artifact, including findings, evidence tables, and routing rules. Use when loaded by squad review author, review critic, or squad lead, not for ordinary code review.
Use when a change alters externally observable behavior or invariant constraints and `docs/system-spec.md` must be updated.
Produces structured incident or bug investigations by tracing code paths, checking recent changes, and using runtime evidence when the prompt provides a data source or commands.