Claude Skills by s-smits

github.com/s-smits

36 skillsA× 360 installs5 views

Attribution And ProofAUse after an Anabasis run, comparison, or system change and before claiming improvement. Attributes movement to one owner, reports identities and censored denominators, separates deterministic proof from model judgement, keeps measurement validity apart from observed exploitation, and states what remains unrun or provisional.researchgitbackendVotes: 0GitHub stars: 61
Bounded InvestigationAUse for 1-7 independent read-only sessions investigating one bounded Anabasis failure, diff, or design question before a fix. Defines non-leading prompts, evidence packets, symmetric fault hypotheses, self-falsification, source adjudication, and concise synthesis. Do not use for whole-run coverage at any session count; use whole-run-investigation instead.researchrustVotes: 0GitHub stars: 61
Codex Luna SwarmALaunch, start, monitor, and collect independent Codex subagents (gpt-6-luna at high, xhigh or max; gpt-6.1-sol for small review batches) for bounded parallel work, from Codex or from Claude Code, including requests supplied as a Markdown file of session prompts. This skill owns Codex subagent transport even when another investigation or review skill defines the questions. Use whenever the user asks for Luna or Sol agents, a swarm, many sessions, a concurrency test, a particular reasoning effo...researchjavascriptgoVotes: 0GitHub stars: 61
Launch RunALaunch authorised Anabasis runs or model pairs through one Bun TypeScript command, or stop an identified run under existing user authority. Owns preparation, exact condition checks, startup and stop mechanics; whole-run-investigation's outcome reference assesses whether stopping is justified.toolstypescriptgoVotes: 0GitHub stars: 61
Oracle HandoverAHand a question to an outside Oracle model (GPT-6 Pro or whichever the operator names) as one task document plus one zip. The Oracle reads GitHub and the harness-builder-v4 history; the zip carries what it cannot fetch: run traces via zip-run, patches for unpublished commits, notes, ledgers, predictions and plans. Use for \"oracle prompt\", \"prompt for GPT-6 Pro\", \"gpt-6-pro appender\", \"zip the traces for the oracle\", or the retired gpt-5-pro-appender.researchgogitVotes: 0GitHub stars: 61
Oss Verifier GroundingASelect, install, and name the authoritative compiler, simulator, solver, or other external verifier for an Anabasis harness. Use when deciding which practitioner tool to trust, writing the truth check that runs it, or diagnosing verifierUnavailable, timeout, crash or sandbox results. Also owns the install probe, the executable board for what the Builder can install and run behind the real cells and wall. Use harness-contract (discrimination proof) for semantic accept/reject coverage.businessrustgoVotes: 0GitHub stars: 61
Prompt Surface CensusAGenerate and tune an AST-derived census of model-visible text in TypeScript or JavaScript. Use to map prompts, hooks, tool descriptions and feedback with their selecting conditions, verify delivery boundaries, and investigate prompt simplification. The census maps source candidates; recorded requests establish actual delivery.developmentjavascripttypescriptVotes: 0GitHub stars: 61
Run Improvement CampaignARun the improvement loop, led by the solve-wall share and the follow-up of each earned fail, with signal a rare confirming event: find the one link holding the climb, choose one simple domain-agnostic change against it, measure it as a matched arm, freeze a prediction, launch, read recorded bytes, decide the next move. Owns experiment selection and evidence reading; the product controller owns build, measure, climb, rebuild, claim and promotion. Also the Super Loop's Meta Agent: a second sess...researchpythongoVotes: 0GitHub stars: 61
Setup PrASet up, reshape, title, publish, version or land an Anabasis pull request the way the operator wants it: where it goes on the stack, when to regroup a back-and-forth history into a few finished commits, how the body reads, who pushes, and when a version tag is attached. Use when asked to set up or append to a PR, regroup or squash its commits, rewrite its title or body, push it, cut a release for it, or merge a stack.documentationgogitVotes: 0GitHub stars: 61
Simplify PrecisionAMeasure and raise the precision of `bun run simplify`, the simplify census, shape by shape. Harvests every recorded census reading and not-slop answer, replays the current scans over history, has blind judges label a seeded sample, scores each shape with a Wilson interval, and tunes a scan's source until the sites it prints are ones a reader would act on (target 0.8). Use when asked how good the census is, why it flags the wrong things, or to tune a tree scan or a staged catcher.researchgoawsVotes: 0GitHub stars: 61
SimplifyAReview the changed code for reuse, simplification, efficiency and altitude cleanups in this repository, then apply the fixes. Quality only — it does not hunt for bugs; use /code-review for that. The method, its scripts and its cases live in this directory, so the skill works without any user-level skill pack.toolstypescriptgoVotes: 0GitHub stars: 61
System Path SimulationATest an uncertain Anabasis path at the smallest real layer that can decide it: deterministic joins, live authoring and admission, semantic host probes, or controlled model comparisons. Use full-run rehearsals only when the question needs them; not for ordinary run review or a path already proved by its owning test.researchgoshellVotes: 0GitHub stars: 61
Wave AuditACompare one or several recorded source states to see whether changes improved Anabasis's climb. Build a state-by-condition census first; compare adjacent states only when prompt, model, command, project lineage, budget and runtime identities support it; read equal windows, earned failures and the mechanisms that fired. Use for 'did the fixes work', 'is this wave better', or before relaunching on a new source.researchgobackendVotes: 0GitHub stars: 61
Weekly Run ReviewASelect, explain and publish the strongest Anabasis runs from one calendar week. Use for the weekly best-run Automation, top-three/top-five comparisons, and questions about what made a run good, what limited it, or what should happen next. It filters by duration, reads every run through the controller's strict evidence and the outcome readers, reads published main syntheses, and challenges finalists with Luna xhigh.researchgogitVotes: 0GitHub stars: 61
Whole Run InvestigationAInvestigate a live, stalled or completed Anabasis run and turn findings into an evidence-bound fix proposal. Also answers whether a campaign is climbing: how the climb is going, whether the batteries are getting harder, why a difficulty decision keeps repeating, how to climb faster, why the controller chose its action. Also the narrow read: reviewing one campaign run or recorded case through its outcomes, whether a live run is still producing useful evidence. Reads evidence; does not launch t...researchrustgoVotes: 0GitHub stars: 61
WayfinderAPlan a huge chunk of work (more than one agent session can hold) as a shared map of decision tickets on your issue tracker, and resolve them one at a time until the way to the destination is clear.documentationgoreactVotes: 0GitHub stars: 61
Anabasis PipelineAUse when deciding which Anabasis stage owns a change, when carrying one bounded build/measure slice from the user's prompt to evidence, or when deciding what may enter a run: the exact one-line prompt, context files, public catalogues, research and solve-side public data. Maps the product loop, the input contract, the authoring sessions, the adoption gates, the four owners, and validate-or-reopen behaviour. Not a long-running campaign controller.researchgoexpressVotes: 0GitHub stars: 57
Builder Blocking LoopAGet told the moment a live Builder session is blocked (five refused submits in a row, a repeated findings set, twelve checks without acceptance, two environment previews in a row, two hours since a submit or rehearsal last returned) and fix the owner iteratively: read, fix on the owning PR, then keep the run or kill it and relaunch on the upgraded system, and rearm. Use while a paid run is live and the operator wants blocking caught and repaired, not only reported.researchgobashVotes: 0GitHub stars: 57
Commit RAThe commit-R protocol — test-driven rewrite of one compartment. Commit, remove the compartment's tests ("R:" commit), rewrite them from a stated hypothesis so production fails on purpose, remove the production files ("R:" commit), then rewrite production and patch or delete the outdated adjacent functions. Use when the operator says "commit R", "commit-R protocol", "remove and rewrite the tests", or asks for a TDD pass that looks for a faster, simpler path to the same outcome.researchgotestingVotes: 0GitHub stars: 57
Final Harness AuditAReview every harness version the Builder produced for one domain. Join workspace commits to iteration evidence, saved execution versions, measured batteries, and claims. Read-only over controller output.toolsbashgitVotes: 0GitHub stars: 57
Harness ContractADesign, change or audit one Anabasis contract: adoption gates, verifier discrimination, artifact representation and the tools over it (including whether a tool earns its place from recorded use), model-visible text, or coding discipline including a small rule-changing fix. Load only the relevant area; source and AGENTS.md own current mechanisms and authority.businessrustVotes: 0GitHub stars: 57
Harness QueryAInvoke an already-built Built Harness on one or more queries, without building a new one. Use when the operator points at a domain bundle and asks to run a task, a family, or a query they wrote, through it.toolsgoapiVotes: 0GitHub stars: 57
Intelligent RebaseACheck semantic interference when concurrent changes or stacked PRs meet. Use after a merge or rebase, at a stack checkpoint, or when comparing changed source with a live run. Measure overlap and exercise the affected combined path; stack-hop owns publication and failed-checkpoint isolation.ai-agentsjavascriptpythonVotes: 0GitHub stars: 57
Jev SurfacesAFind the places in Anabasis where TypeSafe's Jev, a cheap calibrated System One classifier (Noul, Choice, Score over a text state), could replace or sit beside an LLM call, an embedding classifier, a prose regex, a lexical ranking or a blind-judge labelling loop, and prove each one in shadow against recorded labels before anything moves. Sweeps the tree for the five shapes, puts every candidate through the authority filter (what owns the decision, can it touch a pass, a score, a denominator o...researchgoexpressVotes: 0GitHub stars: 57
Main JudgeAUse when adding, changing or reviewing Anabasis's Main Judge: the battery review, its recorded disagreements and the advice they feed. Keeps the Judge outside verification, evidence-bound, separately reported, and unable to alter pass, acceptance, claimability or adoption.researchrustgoVotes: 0GitHub stars: 57
Model Condition ComparisonAUse when the operator wants two or more model conditions (Opus 5, Fable 5.1, Sol) compared on one source head and one prompt, or wants a harness change that must serve more than one model. Covers the condition design that separates harness quality from solver capability, the recorded-evidence join, the reading of each bucket, and the rule under which a change is admitted for every model rather than one.researchexpressbackendVotes: 0GitHub stars: 57
Monitor Session Until IdleACritically supervise another Codex or agent session without taking it over, or audit a long, compacted or delegated session for commitments it forgot, narrowed or left unverified. Maintain executable closure, material-claim, and intervention ledgers; challenge unsupported claims before decision boundaries; correct proved errors; and stop only after accepted work and owned processes close. Use for long-running tasks, campaign watches, delegated reviews, and likely-mistake checks.toolsgoshellVotes: 0GitHub stars: 57
Reduce ComplexityAMeasure whether a change made the stack simpler, on numbers rather than impressions — line counts, branching, coupling, interface width, rule count and model-visible prose — and compare revisions or a whole PR stack on all of them at once. Use before claiming a pass reduced complexity, and when deciding which of several PRs added the machinery.ai-agentspythongitVotes: 0GitHub stars: 57
SafeguardsADesign, add, read or retire an Anabasis runtime safeguard: a bounded log-only sensor around a measured brittle decision. Use it for a post-run finding, not as a substitute for the owning product fix, verifier, score or launch gate.researchgoVotes: 0GitHub stars: 57
Skill MaintenanceAAudit and consolidate the Anabasis skill set. Compare current source, observed skill use and callers; move useful guidance to an existing owner and retire redundant entry points. Preserve main-versus-stack differences, scripts, independent review coverage and resolving links.researchgoshellVotes: 0GitHub stars: 57
Stack HopAMove safely to the current stack head, append a stacked PR, or repair an authorised PR chain. Preserve local work, replay each child onto its actual parent as a linear range, publish batches through one push that gates every commit, and isolate a failed checkpoint. A context hop alone does not authorise remote stack repair.toolsshellgitVotes: 0GitHub stars: 57
Standardise Repo PatternsACompare recurring implementation patterns, choose the best existing approach, teach it in the owning skill, and enforce reliable distinctions through existing lint tooling. Use when asked to generalise inconsistent code or turn a repeated anti-slop finding into a convention and fix its remaining occurrences.researchrusttestingVotes: 0GitHub stars: 57
Test AuditAInvoke whenever writing, changing, reviewing, or sweeping tests in Anabasis. A write-time gate for every new or changed test, and a read-first audit that prunes or consolidates tests re-asserting source, duplicating a stronger owner, coupling to implementation, or keeping a test-only export alive. Knows the suite's helpers, fakes, walls, rule-4 invariance tests and runner, and hands removal claims that rest on overlap to test-impact-and-consolidation.testinggobashVotes: 0GitHub stars: 57
Test Impact And ConsolidationAMeasure which test files carry the least evidence and decide which may be merged or removed. Use when the suite has grown, when a run is slow, when asked which tests matter, or before adding a test that may already exist. Ranks by marginal coverage, then adjudicates each shortlisted file by planting faults.testingrustgoVotes: 0GitHub stars: 57
Wall And Bundle IntegrityAUse when designing, reviewing, or reporting Anabasis's three isolation layers, candidate validation, fingerprint, and immutable bundle identity. Covers executed refusal probes, combined access rules, candidate workspace access rules, typed public projection, controller-held authority, and honest disclosure of any boundary that is still contractual.researchrustbackendVotes: 0GitHub stars: 57
Zip RunAZip one recorded Harness Builder run for handover at a chosen depth (--light, --medium, --verbose). Use when the operator asks to zip, bundle, archive or send a run, its traces, its logs or "everything the model saw", or names a size cap. One script, three cumulative levels, a README and MANIFEST inside every zip.ai-agentsgosqlVotes: 0GitHub stars: 57