Skill Claude Code
Find, pick up, implement, or execute a referenced task, ticket, or issue end to end using the project's tracker, source, verification, and delivery rules. Open or update a verified, reviewed PR — or return an evidenced blocker.
Skill Claude CodeCodex
Pricing, limits, and margin gates for a product whose cost of goods is model inference — compute cost per action, cost per active user, and the breakeven limit before choosing a price, then encode the resulting caps as testable invariants. Use when pricing an AI feature or subscription tier, setting rate/usage limits…
Skill Claude Code
Clarify an ambiguous feature or design through direct questions before work is queued. Resolve consequential decisions and preserve only durable intent.
Skill Claude Code
Turn an agreed outcome into tracker issues with acceptance, dependencies, references, and verification. Keep planning proportional to the work.
Skill Claude Code
Forces the laziest solution that actually works; use on coding tasks or when the user says ponytail, be lazy, simplest solution, YAGNI, do less, or complains about over-engineering, bloat, boilerplate, or unnecessary dependencies. Do not use for non-coding requests.
Skill Claude CodeCodex
GitHub Actions PR review bot with Greptile parity — confidence 0-5, risk gates, checks annotations, and fix loop. Install via setup-vskills.
Skill Claude CodeCodex
Commit and push a verified handoff plus its artifacts, only under explicit authority, and prove the push happened by reading the remote SHA back. Use as the final delivery step after artifacts are ready, when the user asks to push work, or when a handoff needs to reach the remote. Refuses to claim success without…
Skill Claude Code
Turn a repo's markdown docs folder into a searchable retrieval graph — router, generated indexes, state file — and open it as an Obsidian vault.
Skill Claude CodeCodex needs its repo
Sets up this skills repo on a new machine — installs the skills with the vskills CLI, configures pstack model routing for any harness, ensures a verification skill, then regenerates the local-only context docs (CONTEXT.md, docs/) that are deliberately not published in the public repo.
Skill Claude CodeCodex
Audit a whole codebase rather than a single diff — establish the boundary and baseline, sweep layer by layer (entry points, domain, data, integrations, auth, tests, ops), check every stated invariant for a real enforcement point, a failing-on-violation test, and a production signal, rank findings by blast radius, and…
Skill Claude CodeCodex
Deliver a ticket whose delivery path is decided by something other than the code — a token/spend cap, a live data migration, restricted git authority, an unwritable tracker, or a pipeline state that is lying about progress. Use when budget is capped, when a change touches production data or costs money to run, when…
Skill Claude CodeCodex
Strict one-ticket TDD for implementation work — write the failing test first at the real production seam, lock the Verification-command before editing, drive red-green at the highest meaningful seam production traffic actually uses, and prove the fix and the provider failure modes (duplicate and out-of-order delivery…
Skill Claude CodeCodex
Fan work out to subagents safely — delegate independent packages to workers without losing ownership: draw provably disjoint lanes, write the full lane brief, order dependent tickets into waves, and keep the parent as the sole owner-gatekeeper that re-runs every gate and performs every commit and tracker write. Use…
Skill Claude Code
User-invoked benchmark for a selected Pi provider/model — measures how fast the model generates tokens: TTFT, decode TPS, average TPS, latency, token usage, and aggregate percentiles. Use when measuring model speed or throughput before choosing it, or when the user invokes test-tps.
Skill Cursor needs its repo
Prove the vskills repository by linting every skill's frontmatter and relative cross-references, then running the test suite. Use before claiming this repo works, after a skill edit, or when a pull request needs verification evidence.
Skill Claude Code
Drive open PRs to merged — re-establish each PR's real state, resolve review threads, conflicts, failing checks, and CI failures under a bounded repair budget, land only the contiguous verified run bottom-up through the project's forge, watch until the checks pass and each merge lands, and stop at the ceiling. Use…
Skill Claude CodeCodex
Bound every repair loop — retrying an identical failure the same way burns budget without generating information. Set a repair budget before fixing, change the hypothesis or the observation when a retry fails the same way twice, and escalate to a blocker report when the budget expires instead of weakening acceptance.…
Skill Claude CodeCodex
When a report and the repository disagree — a worker claims success with no commit, a tracker says done but the gate is red — the repo is right and the report is narrative. Apply whenever consuming status from tickets, workers, comments, dashboards, or pipeline state.
Skill Claude CodeCodex
Only a verdict from a context that did not write the code counts — an author checks what it meant to build, not what it built, and an author-verified claim on its own diff is not enough. Apply before trusting any review, CI-green diff, self-certification, or self-verified work on a nontrivial change, including…
Skill Claude CodeCodex
List and name every delivery constraint — budget, spend authority, git authority, tracker access, time, state trust — before starting work, because an unnamed constraint gets violated and a named one chooses the path. Apply before executing any task whose environment may restrict the obvious path.
Skill Claude CodeCodex
Verify every claimed outcome against the real artifact — run the feature, read the actual value, inspect the diff — never against a proxy, a self-report, or a build pass. Apply after completing any task, before declaring done, and whenever reading evidence produced by a worker, a ticket, or a pipeline state.
Skill Claude CodeCodex
Root-cause a defect from observable evidence — reproduce it first, isolate the variable that decides the failure, fix at the cause rather than the symptom, and prove the fix with the failing case itself. Use to find the cause of a bug, a regression, a crash, a flaky or flaking test, a UI drift or glitch, an idle-CPU…
Skill Claude CodeCodex
Answer a read-only question about the codebase — a walk-through of how a subsystem, queue, endpoint, webhook, or request flow works, where a behavior comes from, what a change could break — from source and runtime evidence, returning file pointers and an honest confidence level instead of a prose dump. Use for "how…
Skill Claude Code
Pick up in-flight work left by a prior session or another agent — re-establish the real state from the snapshot, branch, and tracker, verify which claims still hold, and continue from the proven position rather than the reported one. Use when resuming paused work, taking over a task another agent started, or when a…
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: