← all publishers

ryan-brosas

@ryan-brosas source repo

362 published skills · page 4 of 4

  1. PDF Extract · ryan-brosas
    Use when extracting text, images, tables, or metadata from PDF files. MUST load to choose the correct extraction library based on PDF complexity, simple text vs structured data vs complex layouts.
    0
    installs
  2. Veda Worker · ryan-brosas
    Use when you want a full plan, implement, verify cycle delegated to the Veda worker agent: orchestrate and author the plan yourself, hand the whole design to one worker run that implements with write access, read its report, then run the reviewer. You never implement; branch on report status and escalate after capped retries.
    0
    installs
  3. Code Cleanup · ryan-brosas
    Use when behavior works but the diff is noisy, repetitive, over-complicated, or AI-shaped - lock behavior first, simplify the changed code, and re-verify without expanding scope.
    0
    installs
  4. Paper Design · ryan-brosas bundle
    Use when working with Paper's documented features, themes and CSS-variable tokens, Figma or HTML paste, SVG editing, MCP setup, Snapshot, or checking new Paper capabilities. Supplies platform mechanics; pencil owns exact Figma transfer and paper-component-consistency owns reusable composition.
    0
    installs
  5. AI Foundation · ryan-brosas bundle
    Use when building LLM abstractions or chat/agent UIs: agent step loops with approval replay, typed output envelopes, history round-tripping, tool-drift fingerprints, stream shims, useChat/useObject hook state machines, RSC streamable value/UI protocols, and schema normalization.
    0
    installs
  6. Brainstorming · ryan-brosas
    Use when a rough idea needs clarification before implementation, resolve ambiguous direction by grounding in what exists, framing the problem, exploring real alternatives, and deciding.
    0
    installs
  7. Inspo Qualify · ryan-brosas
    Use when an active project needs external GitHub prior art: assess one candidate from relevant source; acquire a checkout or create an index only when useful and explicitly authorized.
    0
    installs
  8. Pixel Perfect · ryan-brosas bundle
    Use when a design copy must match its source pixel-perfectly — Paper↔Figma fidelity checks, fixing 'it looks off' or 'not accurate' feedback, verifying fills, radii, shadows, type metrics, or fonts against the design source, or before calling a design match done.
    0
    installs
  9. Skill Catalog · ryan-brosas
    Use when useful expertise is missing from the visible skills, a task could benefit from a source-specific foundation, or the user asks what capabilities exist. Discover cold context with native file search; no catalog command required.
    0
    installs
  10. Web Reference · ryan-brosas bundle
    Use when a live website or web page should be captured and studied as frontend, visual, layout, interaction, or design-system prior art for implementation.
    0
    installs
  11. Cap Foundation · ryan-brosas bundle
    Use when building a screen-recording pipeline: multi-track studio recording actor (pause/resume segmented), crash-safe fragmented MP4 recovery with manifest ladders, browser MediaRecorder spooling, streaming multipart upload with uncertain-completion semantics, getDisplayMedia retry ladders, and VFR-safe duration/sync invariants.
    0
    installs
  12. Dub Foundation · ryan-brosas bundle
    Use when building background-job dispatch on a hosted queue, webhook delivery with failure escalation, short-link redirect middleware (geo/device/AB/bot), bulk CRUD APIs that preserve order and report per-item errors, API error envelopes, cursor+offset pagination, policy rate limiting, Redis locks, or multi-rail partner payout processing (Stripe Connect/stablecoin/PayPal/gift-card) with commission aggregation, referral rewards, and partner postbacks.
    0
    installs
  13. Ell Foundation · ryan-brosas bundle
    Use when porting ell (language model programming) — LMP decorators, _lstr origin tracing, lexical-closure versioning, provider SPI, SQL/alembic store, evaluations, studio; prompt-to-program frameworks that version prompts as code (lexical closures + content hashes), trace which model call produced which text fragment through arbitrary string mutations, adapt one call pipeline to many vendor APIs behind a validation-enforced provider interface, or persist/version LLM programs and invocations in SQLite/Postgres with alembic. Source code and direct tests are ground truth; references carry decisive excerpts and graph retrieval.
    0
    installs
  14. Roo Foundation · ryan-brosas bundle
    Use when building an agent harness's safety layer AND context plane: pure approval classification with timeout auto-answer, budget-window spend caps, tool repetition breakers, model-driven persona routing, promise-gated late config, stack-driven request loops, shadow-git checkpoints, non-destructive condense/truncate history management, dual-log persistence with rewind reconciliation, execution-time tool gating, native tool-call stream parsing with fail-fast validation, fuzzy SEARCH/REPLACE diff application, .rooignore access control, mention-to-context pipelines, multi-instance task history stores, budgeted prompt-cache breakpoint placement, bounded ripgrep search transport, tree-sitter definition-outline extraction over a lazy WASM grammar loader, AND a whole-file-mined workspace→Qdrant code-index RAG pipeline (scanner/parser/watcher/orchestrator, 8-embedder batching kernel, pathSegments filter grammar, in-collection indexing markers, dimension-mismatch recreation).
    0
    installs
  15. Ufo Foundation · ryan-brosas bundle
    Use when porting Microsoft UFO² constellation machinery — async DAG task orchestration with a ready-task claim loop, two-copy agent/orchestrator state merging with advancement-priority ladders, fail-open modification waits, dual-flag cancellation ladders, error-funneling task execution with event publication, Kahn's-algorithm DAG validation, dependency-condition promotion of newly ready tasks, status-dispatched agent state-machine rounds, fire-and-forget orchestration with typed completion-event drains, merged-base editing, fail-fast device-assignment validation, busy-device queue contracts, and device-scoped pending-task disconnect cancellation.
    0
    installs
  16. Veda Deep Plan · ryan-brosas
    Use when a single planning call is not enough, architectural design, subtle bugs with no obvious cause, decisions with no clear answer: plan the hardest problems with Veda Deep Thinking where parallel solvers, a judge, and a verifier converge on the best plan. Drives `veda -S deep-TASKNAME -m flash deep`; does not execute.
    0
    installs
  17. Writing Skills · ryan-brosas bundle
    Use when authoring, editing, or verifying a SKILL.md, or promoting a demonstrated procedure into a skill; select evidence of task lift and preserve the catalog contract.
    0
    installs
  18. Zep Foundation · ryan-brosas bundle
    Use when porting Zep's bulk graph-ingestion engine: streaming pipelines, batch/sequential submission, retry safety, entity canonicalization, or memory-backfill validation.
    0
    installs
  19. Agno Foundation · ryan-brosas bundle
    Use when building or porting an agent framework runtime — the single-agent run loop with retries/cancellation/HITL pauses/approvals and conversation time-travel (continue/fork/regenerate), model provider fallback and retry-with-guidance, OR the multi-agent layer: a supervisor/team-leader loop over member agents, an autonomous shared-task-list mode, tool routing for the leader model, cooperative run cancellation with cancel-before-start support, and HITL pause propagation from members to teams. Source code and direct tests are ground truth; references carry decisive excerpts and graph retrieval.
    0
    installs
  20. Code Discipline · ryan-brosas
    Use when implementing, reviewing, or committing code, general code-discipline principles (scope, verification, type-safety, tests, context-gathering) farmed from high-quality repos.
    0
    installs
  21. Codebase Memory · ryan-brosas
    Use when navigating, indexing, tracing, or comparing local and inspiration repositories through the Codebase Memory MCP knowledge graph.
    0
    installs
  22. Codex Websearch · ryan-brosas
    Use when you need current web facts, documentation discovery, or cited search results through the installed Codex web-search extension.
    0
    installs
  23. Dify Foundation · ryan-brosas bundle
    Use when porting Dify's workflow app-execution spine — long-running LLM workflow execution with cancellation, pause/resume, event streaming to clients, or engine-layer persistence hooks. Source code and direct tests are ground truth; references carry decisive excerpts and graph retrieval.
    0
    installs
  24. Evidence Router · ryan-brosas bundle
    Use when choosing an unfamiliar external evidence source and project source or references are insufficient — cold capability map for MCPs, references, and web.
    0
    installs
  25. Grill With Docs · ryan-brosas bundle
    Use when user wants to stress-test a plan against their project's language and documented decisions.
    0
    installs
  26. Htmx Foundation · ryan-brosas bundle
    Use when porting or embedding a hypermedia-driven AJAX/DOM kernel: attribute-inheritance ladder, trigger-spec grammar and dispatch FSM, extended selector engine, priority FormData collection with live proxy, request lifecycle with hx-sync queueing, response-handling table with HX-* control headers, swap/settle choreography with OOB + preserve pantry, hash-gated node processing, history cache LRU, boost/security gates.
    0
    installs
  27. Isso Foundation · ryan-brosas bundle
    Use when porting Isso's self-hosted comment-server internals: SQLite comment/thread storage, moderation modes, guard rate limits, signed edit/moderation tokens, notification fanout, sanitizer pipeline, or the multi-mixin WSGI deployment core.
    0
    installs
  28. Mem0 Foundation · ryan-brosas bundle
    Use when building agent memory: memory add/extract/update/delete, scoped retrieval with metadata filters and reranking, vector-store backends, and SQLite history.
    0
    installs
  29. Mike Foundation · ryan-brosas bundle
    Use when porting MikeOSS legal-AI platform contracts: nonce-fenced prompt-injection spotlighting, server-side quote verification, HMAC download tokens, shared-project access ladders, or SSE tool-stream choreography.
    0
    installs
  30. Nest Foundation · ryan-brosas bundle
    Use when building or porting a DI container, module system, provider lifecycle, bootstrap pipeline, request-scoped injection, route registration/URL composition, validation pipes, or a boot log buffer — NestJS's core injector + router + pipe kernel as source-confirmed contracts.
    0
    installs
  31. Practices To CI · ryan-brosas
    Use when a recurring, objective failure may deserve an automated check; reuse existing project gates before adding a custom validator or CI step.
    0
    installs
  32. Test Generation · ryan-brosas bundle
    Use when writing, expanding, or auditing tests, converting exact recurring failures into gates, or deciding what to test. Catch-first method: prove a test can catch when reproducible, expand the right existing test, and avoid duplicate coverage.
    0
    installs
  33. Veda Foundation · ryan-brosas bundle
    Use when building a coding-agent oracle/verifier: factored verification checks, model/backend resolution with aliases and personas, parallel ensemble retries, resumable checkpoints, context slicing/token budgeting, worker-report handoff protocol, and rating-grade judging stats.
    0
    installs
  34. Visx Foundation · ryan-brosas bundle
    Use when porting or reimplementing visualization primitives from visx (@visx/*) — scale operator pipeline, chart keyboard-nav state machine + ARIA tree, brush extent FSM, zoom matrices, responsive/text/spring primitives: d3-scale config application with correct operator ordering, keyboard-navigable + screen-reader-accessible charts, pixel-to-domain brush selection with ordinal support, cursor-anchored zoom transforms, parent-size responsiveness without layout loops, SVG text measurement/wrapping, animated line enter/exit trajectories, and referentially-stable data/accessor plumbing.
    0
    installs
  35. Waku Foundation · ryan-brosas bundle
    Use when consulting pinned Waku evidence for signed archive staging, extraction validation order and cleanup ownership.
    0
    installs
  36. Aider Foundation · ryan-brosas bundle
    Use when building AI pair-programming harnesses: token-bounded repository context, repairable edits, explicit file and consent gates, plan-to-edit handoff, diagnostic feedback, provider policy, and scoped Git safety.
    0
    installs
  37. Arbor Foundation · ryan-brosas bundle
    Use when explicitly consulting pinned Arbor evidence for terminal input, follow-output, bounded scrollback eviction and alternate-screen history ownership.
    0
    installs
  38. Biome Foundation · ryan-brosas bundle
    Use when building linters/formatters/language tooling: lossless CSTs with green/red splits, flat formatter IRs with best-fitting printing, rule engines with lazy diagnostics and positional suppressions, and event-stream parser drivers with buffered lookahead, speculative disambiguation, and marker-surgery recovery.
    0
    installs
  39. Bruno Foundation · ryan-brosas bundle
    Use when building API-client cores: TLS agent caching with LRU eviction, OAuth2 grant ladders + encrypted token vaults, PAC/system proxy resolution, cookie-jar wrappers, WebSocket connection FSMs, sandboxed expression interpolation.
    0
    installs
  40. Camel Foundation · ryan-brosas bundle
    Use when porting CAMEL-AI multi-agent workforce orchestration foundation — supervisor decomposition, task bus, coordinator routing, agent pooling, quality gates.
    0
    installs
  41. Cline Foundation · ryan-brosas bundle
    Use when porting Cline's agentic-runtime core — context compaction (trigger budget, overflow-recovery ladder, safe cuts, no-LLM fold), compaction state projection with prefix hashes, runtime safety (loop detection, mistake tracker), pending-prompt steer/queue gate, local hub-daemon transport plane (discovery record, mkdir-lock mutex, ensure/retire ladders, WS envelope, subscription refcounts), claim-once env sentinels with supervised-child restart/adoption, hub command router (authority/drain gates, degraded replies), monotonic shutdown coordination, disk-truth cron reconciliation, agenda task persistence kernel (run-admission gates, exactly-once terminals, crash triage, revision+content-hash CAS), strict/tolerant Markdown intent grammars, location containment, todo-tool scope gate, workspace file-index TTL worker, @mention budget matching, ACP stdio bridge (stdio hygiene, fail-closed permissions, streaming/replay, session lifecycle, config options). Source code and direct tests are ground truth.
    0
    installs
  42. Defense In Depth · ryan-brosas
    Use when invalid data causes failures deep in execution, requiring validation at multiple system layers - validates at every layer data passes through to make bugs structurally impossible
    0
    installs
  43. Execution Router · ryan-brosas
    Use when Main alone is insufficient and child, parallel, RLM, actor, or alternate-model execution is already justified — cold escalation reference.
    0
    installs
  44. Flask Foundation · ryan-brosas bundle
    Use when porting Flask's micro-framework (WSGI request lifecycle) — blueprint registration, session/cookie machinery, config loading, or context/proxy architecture — into another framework or a WSGI-compatible reimplementation. Source code and direct tests are ground truth; references carry decisive excerpts and graph retrieval.
    0
    installs
  45. Ghost Foundation · ryan-brosas bundle
    Use when porting Ghost's outbound webhook dispatch engine (HMAC signing, 410 tombstones, SSRF client selection, plan-limit suppression), scheduled-publishing machinery (JWT-signed schedule URLs, in-memory wake ladder, tolerance-based publish decisions), admin session/API-key auth planes (origin-pinned CSRF, email MFA challenges, kid-addressed JWT verification), or the lazy URL service (per-call URL resolution/generation with derived required-shape inference, NQL filter compatibility stripping, thin-resource degrade reporting, canonical reverse lookup, boot readiness gating). Source code and direct tests are ground truth; references carry decisive excerpts and graph retrieval.
    0
    installs
  46. Goose Foundation · ryan-brosas bundle
    Use when porting Block's goose AI-agent provider machinery — provider trait + collect_stream fold kernel, sanitized/migrating conversation model, audience projection, bounded think-tag stream filter, thinking-effort negotiation, declarative prompt-cache semantics, fail-safe NDJSON request logging, RFC 8628 device flows, typed provider error taxonomy, retry/backoff ladder, transient-only gating, formats/* payload-mapping layer (Anthropic request/SSE fold, OpenAI chat replay coherence + stream triage, Databricks dual-dialect dispatch, Google thought-signature continuity, thinking-config table, OpenAI Responses input mapper + typed-event stream fold, Ollama XML tool-call fallback), declarative provider plane (JSON-defined engine factory, env-expanded base URLs, thin OpenAI-compatible client, completions-prefix composition, tri-effect non-streaming mode, /models failure classification, static-list fallback gate).
    0
    installs
  47. Gpuix Foundation · ryan-brosas bundle
    Use when consulting pinned GPUIX evidence for window-option normalization, queued-event destruction, or render-driven native cleanup.
    0
    installs
  48. Lemmy Foundation · ryan-brosas bundle
    Use when building federation/dispatch engines: per-peer DB-backed activity queues with exactly-once cursors, exponential backoff shared across concurrent senders, modulo-sharded worker fleets, community inbox fan-out maps, Announce wrapping, signed inbound receive gates with dedup, and cross-vendor object round-trips.
    0
    installs
  49. Leverage Capture · ryan-brosas
    Use when capture is explicitly requested, a hard procedure recurs, or an expensive failure or milestone yields a concrete reusable candidate; classify it into code, references, foundations, gates, skills, project notes, or nothing.
    0
    installs
  50. Model Resolution · ryan-brosas
    Use when a task or delegated lane needs a concrete model or backend; inspect live availability, choose the smallest sufficient option, probe uncertainty, and escalate only from observed gaps.
    0
    installs
  51. Notion Workspace · ryan-brosas
    Use when the user needs to search, read, create, update, organize, or synchronize Notion pages and databases, or wants a central workspace or second brain in Notion. Prefer this over driving Notion in a browser.
    0
    installs
  52. Plane Foundation · ryan-brosas bundle
    Use when porting multi-instance CRDT document collaboration (a Hocuspocus/Yjs live server bolted onto an HTTP API backend) or SSRF-safe outbound HTTP with signed webhook delivery (fail-closed IP classification, DNS-rebinding-proof pinned fetches, manual redirect ladders, celery delivery/retry/deactivation). Source code and direct tests are ground truth; references carry decisive excerpts and graph retrieval.
    0
    installs
  53. React Foundation · ryan-brosas bundle
    Use when porting cooperative time-slicing schedulers, priority-ordered task queues, delayed-task timers, macrotask yield/error-resume loops, or lane/priority-bitmap scheduling kernels modeled on React's Scheduler (`packages/scheduler`) and the react-reconciler root-scheduling plane (`ReactFiberRootScheduler`, `ReactFiberWorkLoop` render entry, `ReactFiberLane`). Source code and direct tests are ground truth; references carry decisive excerpts and graph retrieval.
    0
    installs
  54. Solid Foundation · ryan-brosas bundle
    Use when porting or building a fine-grained reactive UI framework: signal/memo kernel, pull-based topological updates, ownership & disposal tree, store proxies with keyed reconciliation, control flow, Suspense ladder, SSR shadow kernel + streaming hydration.
    0
    installs
  55. Storm Foundation · ryan-brosas bundle
    Use when porting STORM-style research pipelines: multi-perspective question asking, citation-grounded writing, reference unification.
    0
    installs
  56. Sweep Foundation · ryan-brosas bundle
    Use when porting GitHub-app webhook routers that fan out to long-running agent jobs: fail-open HMAC signature gates, latest-wins ctypes thread cancellation, per-object coalescing work queues, event/action gate ladders, GHA-autofix attribution chains, single-progress-comment lifecycles, branch/commit/PR assembly, cron PR maintenance, delta-gated progress persistence, mutable-comment header rendering, lexical/vector code-search infrastructure, LLM file-selection budgeting, Jira second-forge dispatch via issue mirroring, FCR tag grammar with COPIED_FROM markers, FCR application loops with lazy tool calls, self-correcting search-and-replace match ladders, FCR pre-validation escape hatches, GHA two-stage planner, repo-parsing chunk corpus builder, lint/parse validation kernel, FCR tag grammar kernel, jsonpatch-diffed chat streaming, stateful suggestion streaming, XML tool-call parse twins, and plan-context assembly. Source/tests ground truth; references carry excerpts.
    0
    installs
  57. Turso Foundation · ryan-brosas bundle
    Use when building a SQLite-compatible storage engine: bi-temporal MVCC with Hekaton-style commit dependencies, checksum-chained WAL framing with three-phase commit, b-tree rebalancing, and pin-count durability.
    0
    installs
  58. Txtai Foundation · ryan-brosas bundle
    Use when porting txtai vector-engine mechanics — hybrid dense+sparse fusion, IVFSparse/FAISS/sqlite-vec index builds, BM25 term scoring, filter-then-search planning, or embeddings lifecycle. Source code and direct tests are ground truth; references carry decisive excerpts and graph retrieval.
    0
    installs
  59. Umami Foundation · ryan-brosas bundle
    Use when porting privacy-first analytics/telemetry machinery — cookieless derived session identity, rolling cache-token handshakes for anonymous ingest, dual-backend SQL dispatch (Postgres + ClickHouse), dynamic filter compilation with typed bind placeholders, Kafka wire-size batching, soft-delete read caches, 2FA with partial-auth tokens and replay ledgers, rrweb session-replay chunking/reassembly, heatmap capture with scroll bucketing, hand-rolled Core Web Vitals, and share-token capability grants.
    0
    installs
  60. Affine Foundation · ryan-brosas bundle
    Use when porting CRDT sync loops, offline-first storage ladders, or block-editor data layers — AFFiNE local-first collaboration stack: Yjs sync engine (DocEngine/SyncPeer), reactive block store (stash/pop, flat Y.Map proxies, schema globs), and production doc-sync job kernel with clock-map dedup.
    0
    installs
  61. Celery Foundation · ryan-brosas bundle
    Use when building a distributed task queue or job worker: at-least-once ack/reject semantics across early/late modes, retry ladders (manual Task.retry, autoretry wrapper, exponential backoff with full jitter), prefork pool with event-loop time limits, bootstep blueprint startup/shutdown graphs, broker reconnect loops, ETA/countdown timers, beat's reentrant heap scheduler with crontab DST math, gossip worker elections, and mailbox-based remote control.
    0
    installs
  62. Chroma Foundation · ryan-brosas bundle
    Use when porting Chroma embedded vector database internals — HNSW segment engine, layered brute-force batches, SQL metadata filter planner, SQLite WAL queue, and the Rust MaxScore/SPANN/HNSW-provider engine.
    0
    installs