Claude Skills by OutlineDriven

Use when a workflow, plan, diff, or completed work cycle must be checked for practices that punish the project later: symptom patching, infinite retries, weak verification, score chasing, budget burn. Returns a read-only evidence report naming each detected practice with a cheaper alternative. Not for source, remote, credential, publish, deploy, or irreversible changes.

Use when a user asks for a gut-check on a decision or action, or asks whether enough is known to proceed. Produces a prose-only assessment with concrete knowns, unknowns, and one verdict: Proceed, Proceed with caveat, Pause, or Blocked. Not for numeric confidence scoring.

Use when a user wants a readable, shareable HTML document of findings. Produces a self-contained HTML readout under ~/.readouts, embedding cited source text and refreshing the local index. Not for remote, credential, publish, deploy, or irreversible changes.

Use when asked to recover prior work and current status before resuming. Returns a status-tagged capsule and one concrete next move. Not for handoff snapshots — use handoff. Not for tasks that require source or remote-system changes.

Use when the user runs /scrape with a URL to extract page data and media through a browser workflow and save assets with a manifest to a local directory. Read-only; not for any mutating page action.

Use when the user asks to beam, publish, or share the current local coding session to an authenticated remote receiver. Builds a redacted, size-bounded JSON payload from visible messages and aggregate tool counts, previews it for human confirmation, then POSTs it to a configured endpoint and returns the shareable URL. Not for viewing a transcript locally; use session-viewer.

Use when the user asks to view, export, or inspect a session transcript in a browser. Produces one local single-file searchable HTML viewer from the session JSONL with credential scrubbing and optional browser launch. Not for sharing a session — use session-share.

Use when the user wants a per-finding visual walk through a diff or PR: one finding per turn with a Keep/Skip/Discuss choice and a close tally. Not for written review reports — use review or pr-review; not for codebase tours — use show-me.

Use when asked for a research-backed answer requiring codebase and documentation investigation, or when answering requires reading many files or conducting a wide survey. Investigates repository questions with sourced evidence, spawning parallel research subagents when the scope is wide. Not for file edits, diffs, or patches; use fix. Not for remote-system changes.

Use when the user asks for a weekly sentiment report, weekly social summary, or how mentions looked this week. Compares the most recent completed week against the prior week: sentiment scores, volume deltas, notable patterns, and product feedback, published as a PR after human approval. Not for continuous monitoring or alerting.

Use when asked to find duplication, check consistency, establish or repair SSOT, consolidate scattered facts, or when the user says "consolidate this" or "ssotize this". Audits every occurrence, proposes a reversible plan, then folds unique details into one approved canonical home. Not for remote or irreversible changes.

Use when the user asks to summarize a set of documents, identify recurring themes and conflicting statements across multiple files, or synthesize internal docs. Reads a provided document set, extracts cited themes and conflicts, and writes a report. Not for single-document summaries or non-text sources.

Use when the user wants ongoing teaching across sessions in a persistent workspace, or when multi-session learning needs a sourced workspace where every lesson cites a trusted source and advancement requires demonstrated retention. Builds a mission, resources, lessons, learning records, glossary, and shared assets. Not for one-off explanations; use explain-concept.

Use when the user wants an async questionnaire for someone else, a discovery questionnaire, or a knowledge gap needs answers from outside the repository. Produces prioritised one-idea questions with answer stubs and ambiguity rationale. Not for direct conversation elicitation; use askme. Not for agent-driven research; use research. Not for remote, credential, publish, deploy, or irreversible changes.

Use when a load-bearing claim is unverified, a plausible statement has never been checked, the user says "fact-check this" or "verify this claim", or an artifact or sentence about to be written asserts something as plausible, absurd, novel, or impossible and doubt arises. Tests every claim in both directions (could the absurd be real? could the obvious be false?), cites sources, and returns a verdict per claim. Standalone-claim mode is read-only; artifact mode also corrects mechanically-clear...

Use when a measurable claim needs before/after proof. Runs controlled baseline and treatment probes, then returns VERIFIED, NOT VERIFIED, or INCONCLUSIVE with deltas. Not for fact-checking or done-claim gating — use verify-both-ways or verification-before-completion.

Use when asked to find distribution opportunities for a template, tool, or artifact. Classifies the artifact, queries known directories and aggregator platforms, filters for active communities, scores reach and fit, and compiles a ranked report with submission rationales. Not for content creation or campaign management.

Use when a developer returns to a project or loses context. Produces an 8-section HTML recap page so they can rebuild the mental model and derive next steps only from evidence. Not for session handoff snapshots; use handoff.

Use when the user says "wait, what", "the explanation is unclear", "say that again", or asks to restate the last response in plain language. Re-pitches the previous explanation with the missing frame supplied, in plain English and the project''s ubiquitous language. Don''t use for tasks that require source or remote-system changes.

Use when the user wants to be walked through code rather than handed a report: "walk through this", "guided code walk", or "explain this codebase". Renders a visual shape and hands the user the next step each turn. Don''t use for tasks that require source or remote-system changes.

Use when asked to summarize authored work from the last week. Reads git history without changing it, then returns an executive summary and work classification. Don''t use for tasks that require source or remote-system changes.

Use when a user asks what was completed in a named date or commit range. Reads git log and returns a concise dated status report. Don''t use for tasks that require source or remote-system changes.

Use when the user asks "why does X work this way", requests design rationale, a postmortem, or a data-backed threshold. Dispatches parallel read-only investigators across seven evidence categories and returns a confidence-weighted cited narrative. Not for tasks that require source or remote-system changes; not for current runtime behavior — use how.

Use when a user asks to mine recent chats for workflow preferences. Reads chat history, extracts recurring patterns, and returns an evidence-backed preference synthesis with proposed workflow artifacts. Don''t use for tasks that require source or remote-system changes.

Use when asked to run /artifact-arena to generate and judge competing artifact implementations. Produce one verified synthesis with provenance and rejection reasons. Don''t use for remote, credential, publish, deploy, or irreversible changes.

Use when a human has an approved delivery plan and wants the full chain run under phase gates. Delegates implementation to work, simplification to simplify, review to review, and finalization to review-and-ship. Not for planning, open-ended debugging, or single-step execution — use work directly.

Use when a human explicitly runs /cloud-task-orchestrator for a large task across cloud agents to drain a verified task graph. Not for local subagent coordination — use orchestration-patterns.

Use when a candidate patch or answer needs independent corroboration before it is trusted. Run N isolated sandboxed attempts and return the strict-majority byte-identical patch with its agreement count, or report no consensus. Not for multi-stance investigation — use council.

Use when a configured loop misbehaves, produces unexpected results, or its setup soundness is questioned. Reads the charter, state, gate, and budget files, classifies health as healthy, warning, or blocked, and emits at most three prioritized actions. Not for tasks that require source or remote-system changes; not for loop design — use harness-engineering.

Use when asked to /feedback-sweep [setup|reconfigure] [mode:non-interactive]: sweep configured sources, acknowledge user-approved items, verify fixes merged to default, and emit an autopilot-ready plan. Not for pattern discovery — use feedback-pattern-sweep.

Use when asked to design a bespoke execution workflow when no playbook fits; verify a falsifiable predicate on the real product with a reviewable trail. Don''t use for remote, credential, publish, deploy, or irreversible changes.

Use when asked to evaluate a proposed commit, merge, or auto-merge against a version-1 gate config. Returns an allow or deny verdict naming matched paths; performs no merge. Not for landing PRs — use gate-and-merge. Read-only.

Use when a human runs /merge-and-deploy to merge a PR and trigger or verify deployment. Don''t use for tasks that require source changes or without explicit human confirmation at each gate.

Use when a high-priority run reaches at least 90% of its budget and requests an extension. Ask its human owner once; on grant record the bounded extension, on decline mark the run WAITING_FOR_BUDGET. Never self-grants. Not for low-priority runs — let those exhaust silently.

Use when work decomposes across subagents or role panels: select a coupling-based orchestration mechanism, spawn bounded subagents, verify every artifact owner-side, and write only the reconciled synthesis. Not for cloud-agent task-graph orchestration — use cloud-task-orchestrator.

Use when asked to partition non-overlapping scopes across subagents and drive each to completion with fresh review, fix reports, and a glanceable actual-work map. Not for concurrent independent sub-tasks — use parallel-launch.

Use when asked to apply the pstack rigor mode to non-trivial work. Routes the task to a matched playbook from a 23-index and executes it to verified completion on the real surface. Not for remote, credential, publish, deploy, or irreversible changes.

Use when asked to change state in an external system, propose the write locally and halt at the human gate without executing it. Don''t use for remote, credential, publish, deploy, or irreversible changes.

Use when a human runs /publish-release-pr to ship a versioned, redaction-gated PR with evidence. Not for tag-and-push releases — use release-gate.

Use when a user runs saga or asks to autonomously build a sizable feature. Produces a spec tree, delegates to worker subagents in isolated worktrees, validates at each milestone, and gates user acceptance before VCS commit. Not for executing a given plan — use subagent-driven.

Use when a user says execute with subagents or hands over an ordered multi-task plan to run a per-task implementer-reviewer loop with an audit gate. Don''t use for credential, paid, or deployment work, or for any mutation outside the skill''s designated workspace.

Use when asked to run partitioned parallel coverage or races across isolated workers, producing a consolidated evidence table with gaps and dropouts. Don''t use for remote, credential, publish, deploy, or irreversible changes.

Use when a scheduled or watcher tick fires and a lightweight pulse must capture current external state. Not for remote, credential, publish, deploy, or irreversible changes.

Use when asked to drive decomposable work as a deterministic multi-subagent workflow with phased fan-out and adversarial verification. Also handles audits, migrations, broad research sweeps, or scale one context cannot hold. Don''t use for remote, credential, publish, deploy, or irreversible changes.

Use when a completed agent run must be recorded as durable, queryable evidence. Appends one entry to the JSONL run log under an ISO date guard, prunes expired entries, and updates the last-run pointer. Don''t use for remote, credential, publish, deploy, or irreversible changes.

Use when the user says "atomic PRs" or requests one issue or PR per logical change. Don''t use for single-change pushes or uncommitted change-sets.

Use when the user says release this, publish this package, or cut a release for a changesets-based npm package. Not for non-npm packages or releases without a changesets workflow.

Use when the user runs /browser-cookie-store to populate the session cookie store from installed browsers for authenticated browsing. Discovers browser profiles, reads and decrypts local cookie databases, extracts session cookies, and writes them to the destination store. Not for remote, credential, publish, deploy, or irreversible changes.

Use when the user runs /browser-qa to run the QA verification pass and return report-only results without entering a fix loop. Runs configured build, test, and lint checks in order, collects findings, and writes one local report. Not for remote, credential, publish, deploy, or irreversible changes.

Use when the user asks to cherry-pick, backport, or apply a hotfix to a release branch. Not for merging feature branches or cutting new release branches.