← all publishers

stevennitesh

@stevennitesh source repo

68 published skills

  1. Wizard 2 · stevennitesh bundle
    Create one repository-native interactive script for a bounded, settled procedure with at least two dependent stages only the current human can perform. Use only when explicitly selected and scripting materially reduces secret exposure or configuration mistakes; exclude agent-executable work, single manual actions, unresolved decisions, questionnaires, and ordinary implementation.
    0
    installs
  2. Research 2 · stevennitesh bundle
    Research one bounded source-answerable question for a caller-owned use. Use when claim-owning sources and evidence judgment are needed for a cited answer or one authorized note; exclude simple lookup, open surveys, diagnosis, runnable probes, stakeholder-held gaps, and requests to make the caller's decision.
    0
    installs
  3. Hillclimb 2 · stevennitesh bundle
    Repeatedly improve one measurable runtime, resource, cost, capacity, or product outcome of a fixed representative workload against a settled target. Use only when explicitly selected; exclude one-off fixes, causal diagnosis, disposable design probes, and code-shape simplification.
    0
    installs
  4. Prototype 2 · stevennitesh bundle
    Prototype one bounded design question with a disposable runnable probe; exclude production proof, uncertain defects, and multi-decision design.
    0
    installs
  5. To Tickets 2 · stevennitesh bundle
    Explicitly turn one settled source into the smallest useful dependency-ordered implementation ticket graph and publish it to the configured tracker. Return one bounded item to Implement without creating a graph.
    0
    installs
  6. Wayfinder 2 · stevennitesh bundle
    Chart or advance one finite tracker-backed route from a bounded foggy destination to coherent settled source for a specification.
    0
    installs
  7. Change Review 2 · stevennitesh bundle
    Review one selected branch, WIP, staged or since-X diff, PR, or fixed implementation candidate read-only for concrete correctness and code-quality problems. Use for direct review or caller-triggered independent review; exclude repository-baseline audits, repair requests, and explicit High Assurance Review.
    0
    installs
  8. To Spec 2 · stevennitesh bundle
    Explicitly turn one named settled source into one verified durable parent specification; exclude shaping, ticket slicing, implementation, and delivery.
    0
    installs
  9. Audit Codebase 2 · stevennitesh bundle
    Build or continue an organized HTML map of a repository, audit one user-selected subsystem for architecture and code-quality problems, or analyze one user-selected candidate. Use explicitly for thorough repository-baseline improvement discovery; exclude pending diffs, implementation, ticket publication, and release decisions.
    0
    installs
  10. Repo Bootstrap 2 · stevennitesh bundle
    Use only when the current user explicitly asks to inspect, create, or reconcile one bounded repository's Programming Agent Skills setup; apply only an exact approved delta.
    0
    installs
  11. Codebase Design 2 · stevennitesh bundle
    Design one bounded module or interface when architecture itself is unresolved. Use for an explicit architecture request or a consequential question about ownership, data shape, interface, state or failure policy, seam, or migration; exclude ordinary implementation with a clear repository-native design and codebase-wide discovery.
    0
    installs
  12. Context Hygiene 2 · stevennitesh bundle
    Audit persistent coding-agent context or judge whether a lesson belongs in memory, a current source, project guidance, enforcement, task state, history, or nowhere. Use for memory admission, reflection, cleanup, or pruning; exclude live-thread compaction, ordinary documentation cleanup, and implementation of routed project changes.
    0
    installs
  13. Diagnosing Bugs 2 · stevennitesh bundle
    Explicit-only investigation and optional causal fix for hard, intermittent, performance, environment-only, production-only, or otherwise causally ambiguous bugs.
    0
    installs
  14. Parallel Implement 2 · stevennitesh bundle
    Deliver one explicitly requested fixed set of two or more accepted implementation items with concurrent workers when their ownership and write effects are independent. Root-only; exclude shaping vague work or implementing one item.
    0
    installs
  15. Writing For Agents 2 · stevennitesh bundle
    Create, edit, or audit instructions agents consume, including skills, AGENTS.md, prompts, guides, specs, tickets, handoffs, and subagent assignments used in orchestration. Exclude ordinary human-facing prose, application code, and orchestration mechanics such as worker selection, scheduling, or integration.
    0
    installs
  16. Skill Router 2 · stevennitesh bundle
    Route only an explicit next-skill selection request or a terminal out-of-scope residual packet delegated by another skill. Return one next skill or none without executing it. Do not replace an active skill's in-scope work or owned handoffs.
    0
    installs
  17. Repo Bootstrap 3 · stevennitesh bundle
    Bootstrap or reconcile a target repo's verified setup surface for the Programming Agent Skills pack.
    0
    installs
  18. Codebase Design 3 · stevennitesh bundle
    Design one bounded module behind a small caller-facing interface. Use when choosing the interface, seam, adapter, or caller-facing test surface for one specific module, or when another skill needs shared deep-module vocabulary.
    0
    installs
  19. Resolving Merge Conflicts 2 · stevennitesh bundle
    Inspect or resolve an active conflicted merge, rebase, cherry-pick, revert, or unmerged index. Status, explanation, and review are read-only. Exclude operation planning or start, ordinary diffs, completed operations, marker-like text without Git conflict state, and post-operation bugs.
    0
    installs
  20. Tdd · stevennitesh bundle
    Test-driven development. Use only when the user explicitly requests TDD, test-first work, or RED-GREEN-REFACTOR, or applicable repository policy requires TDD. Requests for tests, integration tests, regression tests, or coverage alone do not trigger it.
    0
    installs
  21. Wizard · stevennitesh bundle
    Create a guided local script for a settled procedure that needs human identity, private input, dashboard access, or physical action. Use when explicitly requested and guidance reduces ordering, target-selection, or configuration mistakes; exclude unresolved design interviews and ordinary agent-executable work.
    0
    installs
  22. Triage · stevennitesh bundle
    Sort raw tracker issues and configured external PR or MR intake into one honest disposition or concise handoff, then apply authorized tracker changes with read-back. Exclude project-created ready work, implementation, deep diagnosis, and code review.
    0
    installs
  23. Research · stevennitesh bundle
    Research questions or compare options requiring source evaluation, synthesis, or reconciliation of conflicting evidence. Exclude quick factual lookups and runnable experiments.
    0
    installs
  24. Handoff · stevennitesh bundle
    Compact the live thread into one verified local continuation packet for a fresh Codex context that can access the same work root.
    0
    installs
  25. To Spec · stevennitesh bundle
    Explicitly turn one settled source into a verified durable parent specification when several slices, multi-session coordination, or a lasting decision boundary justify it. Return one bounded item to Implement when a spec would add no value.
    0
    installs
  26. Hillclimb · stevennitesh bundle
    Iteratively improve a measurable runtime, resource, cost, capacity, or quality outcome through controlled local experiments. Use for a requested optimization loop against a representative workload; exclude causal debugging, one-off feasibility probes, and unmeasured code cleanup.
    0
    installs
  27. Prototype · stevennitesh bundle
    Build a runnable experiment to test design feasibility, behavior, interaction, or performance before committing to an approach. Exclude production implementation and existing-bug diagnosis.
    0
    installs
  28. Grilling · stevennitesh bundle
    Grill one bounded user-owned or approved caller-owned decision until shared understanding is confirmed. Use for direct stress-testing or an invoked conversation-only decision; remain before action.
    0
    installs
  29. Shape Work · stevennitesh bundle
    Clarify or stress-test an idea into an implementable outcome, capture a useful specification, or directly reconcile domain meaning, invariants, context relationships, and settled decisions. Use for shaping work or requested domain-record updates; exclude routine vocabulary lookup, straightforward implementation, and delivery-ticket decomposition.
    0
    installs
  30. To Tickets · stevennitesh bundle
    Draft, publish, or repair delivery tickets from settled work, with acceptance and real dependencies. Use only for explicit ticketing requests.
    0
    installs
  31. Implement · stevennitesh bundle
    Deliver one explicitly selected bounded ready item through a small, sound implementation and proportionate proof.
    0
    installs
  32. Wayfinder · stevennitesh bundle
    Coordinate one bounded destination whose interdependent decision questions need tracker-backed multi-session resolution. Exclude one-session decisions, settled specifications, implementation planning, and delivery.
    0
    installs
  33. Value Stock · stevennitesh bundle
    Research and value a publicly traded company from current primary evidence. Use for fair value, intrinsic value, DCF or reverse DCF, residual income, forward P/E or PEG, comparable-company valuation, earnings or guidance tied to valuation, margin of safety, or a fundamental thesis. Route unsupported methods to explicit capability gaps. Do not use for technical analysis, short-term price forecasts, trade execution, portfolio sizing, or personalized investment advice.
    0
    installs
  34. Tdd Slice · stevennitesh
    Use when approved work changes caller-visible or user-visible behavior, adds a new behavior path, or an understood bug needs a new durable behavior or regression check.
    0
    installs
  35. Mle Workflow · stevennitesh bundle
    Align, build, review, harden, or operate a production or reproducibility-bound machine-learning system across project purpose, data and feature contracts, experiments, evaluation, testing, release, monitoring, retraining, incidents, and retirement. Use for predictive or generative ML, ranking, recommendation, forecasting, anomaly detection, embeddings, batch or online inference, notebook-to-pipeline work, model refreshes, leakage or training-serving skew, promotion decisions, and MLE production-readiness reviews. Exclude one-off exploratory analysis with no reproducibility or operational requirement, algorithm tutorials, and work whose primary challenge is application code rather than ML-system behavior.
    0
    installs
  36. Next Slice · stevennitesh bundle
    Execute one implementation slice from a saved slice plan.
    0
    installs
  37. Slice Plan · stevennitesh
    Use when approved multi-step repo work truly needs a written execution plan or handoff with task boundaries, checks, execution mode, ownership scope, continuity, and final verification; skip tiny edits, one behavior, one bug, or issue-only/PR-only tracking.
    0
    installs
  38. Change Review · stevennitesh bundle
    Review a diff, branch, PR, or uncommitted changes for correctness and maintainability problems. Exclude whole-codebase audits and implementation without a review request.
    0
    installs
  39. Skill Router · stevennitesh bundle
    Choose one next engineering skill when the user explicitly asks which skill to use or invokes $skill-router. Return that skill or none, then stop without starting it.
    0
    installs
  40. Audit Codebase · stevennitesh bundle
    Find and rank evidence-backed improvements to an existing codebase's architecture and maintainability. Optionally create or continue a visual system map and coverage atlas. Use for baseline improvement discovery; exclude reviewing a pending diff and implementing fixes.
    0
    installs
  41. Repo Bootstrap · stevennitesh bundle
    Inspect, create, or reconcile a repository's agent guidance when the user requests setup or a specific guidance repair. Covers repository instructions and engineering conventions; exclude ordinary coding, environment installation, and automatic setup merely because a preferred document is missing.
    0
    installs
  42. Simplify Code · stevennitesh bundle
    Simplify one user-selected existing-code target without changing accepted behavior. Use only when explicitly selected; exclude new behavior, bug fixes, broad discovery, and unsettled interface or ownership decisions.
    0
    installs
  43. Codebase Design · stevennitesh bundle
    Resolve unsettled architecture or feature-integration decisions about ownership, interfaces, state, dependencies, or migration. Exclude routine implementation and broad codebase audits.
    0
    installs
  44. Context Hygiene · stevennitesh bundle
    Assess completed-work lessons or audit persistent agent context for useful knowledge, clear ownership, loading paths, duplication, and staleness. Apply requested cleanup; excludes live-thread compaction and ordinary documentation editing.
    0
    installs
  45. Diagnosing Bugs · stevennitesh bundle
    Root-cause difficult bugs, intermittent failures, or environment and cross-system faults when the cause is unclear. Exclude obvious fixes and optimization without a defect.
    0
    installs
  46. Author Skills · stevennitesh
    Use when creating, consolidating, editing, testing, or reviewing coding-agent skills, skill packs, orchestration skills, reusable software-engineering workflows, or instructions meant to steer coding-agent behavior in repos.
    0
    installs
  47. Clarify Scope · stevennitesh
    Use when cheap repo evidence cannot resolve a material decision about target behavior, public or caller contract, test strategy, data/state/security, release or rollback risk, ownership boundary, or durable decision records before safe implementation can proceed.
    0
    installs
  48. Coding Router · stevennitesh
    Use when starting nontrivial repo work where route choice, scope, assumptions, or completion evidence could drift; defaults to a micro-loop and prevents skill pileups.
    0
    installs
  49. Diagnose Loop · stevennitesh
    Use when tests, commands, builds, CI jobs, or workflows fail; behavior is wrong; output or performance regresses; flakes or crashes appear; or an error is not yet understood.
    0
    installs
  50. Pre Pr Review · stevennitesh
    Use when reviewing a branch, commit, working tree, or PR-ready diff before opening or updating a pull request; when asked for pre-PR review, PR readiness, bug finding, regression review, code review, review before merge, or review with Codex; and when the review should focus on actionable correctness, security/privacy, data, contract, test, dependency/config, performance, or migration issues introduced by the diff.
    0
    installs
  51. To Slice Plan · stevennitesh bundle
    Create no-babysitting implementation slice plans.
    0
    installs
  52. Domain Modeling · stevennitesh bundle
    Resolve or capture project-specific domain meaning, invariants, bounded contexts, relationships, or an already-settled ADR candidate. Exclude vocabulary lookup, code-structure design, and unresolved product or architecture decisions.
    0
    installs
  53. Grill With Docs · stevennitesh bundle
    Use when one user-owned repo-backed decision needs live grilling while its domain meaning, invariants, or relationships are kept current; exclude conversation-only grilling and already-settled domain capture.
    0
    installs
  54. Cost Aware Coding · stevennitesh bundle
    Propose a cost-aware coordination plan or execute an accepted one. Use only when requested for coding work; exclude ordinary implementation and standalone model comparisons.
    0
    installs
  55. To Questionnaire · stevennitesh bundle
    Create one Markdown discovery questionnaire for one external stakeholder who owns facts, judgment, or a decision unavailable from inspectable sources and the user.
    0
    installs
  56. Github Tracking · stevennitesh
    Use when repo work needs durable GitHub recordkeeping: issue/PRD/slice bodies, claim/status/blocker comments, PR/issue links, readiness fields, or CI/review evidence recorded in GitHub; skip ordinary local work, generic triage, PR publishing, CI debugging, and review-comment resolution when narrower workflows own them.
    0
    installs
  57. Repo Onboarding · stevennitesh
    Use when entering an unfamiliar repo, adopting this skill pack in a repo, resuming work when stale commands/instructions/context matter now, or when build/test/GitHub/context conventions are unknown before coding.
    0
    installs
  58. Parallel Implement · stevennitesh bundle
    Coordinate an explicitly requested parallel implementation across bounded items, with exclusive ownership, safe integration, cross-item proof, and recoverable worktree cleanup. Use for implementation fanout, not parallel research or a single coding task.
    0
    installs
  59. Writing For Agents · stevennitesh bundle
    Write or audit agent-facing skills, AGENTS.md, guides, specs, prompts, and continuation handoffs. Exclude product discovery, ordinary prose, minor wording edits, and app task transfers.
    0
    installs
  60. Codebase Cleanup · stevennitesh
    Use when the user asks for cleanup beyond incidental touched-file polish: explicit human-reviewability cleanup for comments/docstrings/imports/code order, cleanup discovery, continued cleanup loops, behavior-preserving refactors with structure risk, duplicated logic, module-boundary or caller-interface simplification, testability blockers, or obsolete-code removal.
    0
    installs
  61. Workspace Safety · stevennitesh
    Use as a lightweight gate when overlapping dirty tracked paths, relevant untracked protected paths, or repo operations could overwrite, delete, publish, mix, stage, regenerate, or depend on unknown work: staging, commit, PR, push, merge, branch/worktree changes, installs, lockfiles, migrations, generators, broad formatters, generated cleanup, or destructive Git/filesystem commands.
    0
    installs
  62. Slice Plan Runner · stevennitesh bundle
    Run a saved slice plan through multiple committed slices.
    0
    installs
  63. Subagent Workflow · stevennitesh bundle
    Use when the user explicitly asks for subagents, parallel agents, worker agents, independent code review, multi-agent codebase exploration, subagent-ready handoffs, or an approved plan/issue calls for authorized, bounded delegation that is cheaper than local work.
    0
    installs
  64. Verify Before Done · stevennitesh
    Use before claiming repo work is done, fixed, passing, reviewed, ready, resolved, clean, safe to merge, mergeable, or complete; calibrates claim strength to fresh evidence.
    0
    installs
  65. Worktree Isolation · stevennitesh
    Use when approved parallel implementation across agents or sessions needs separate git branches/worktrees, when a user explicitly asks for isolated worktrees, or when `parallel-disjoint`/`parallel-overlap` execution needs branch, ownership, integration, or cleanup safety before editing.
    0
    installs
  66. High Assurance Review · stevennitesh bundle
    Thorough read-only review of one fixed code candidate through two fresh independent reviewers and one verified decision. Use only when the user explicitly requests high-assurance, heavy, or final review, commonly for an implemented ticket graph or before merging a PR. Exclude ordinary review, repository-baseline audits, repair, and mutation.
    0
    installs
  67. Issue Driven Execution · stevennitesh
    Use when approved repo work should be coordinated through a plan document, GitHub issues, one verified implementation or research issue at a time, and an explicit checkpoint policy.
    0
    installs
  68. Resolving Merge Conflicts · stevennitesh bundle
    Inspect or resolve conflicts in an active Git merge, rebase, cherry-pick, revert, or unmerged index. Exclude planning integrations and debugging completed merges.
    0
    installs