RetentaPractical architecture guides and source-linked skill records for agents that work across runs. Publisher provenance, source review, runtime testing, and security verification stay separate—no popularity score is presented as trust.
Field guides answer a concrete implementation question and link every adjacent platform claim to an official source. Metadata-only imports never appear here.
A skill gives an agent a procedure. Durable memory preserves what happened, what remains open, and what the agent learned after the current context disappears.
Read the label
Filter 61 original source assessments across 19 publishers. “Official” identifies the source owner; “unverified” means Retenta has not executed or independently security-tested it.
Showing 61 of 61 records.
Retenta · Claude Code · Codex · Gemini CLI
Retenta's flagship skill: an agent that kicks off or resumes a software project, interviews the owner, keeps docs/ and GitHub Issues as one record, and delegates work through approved plans and reviewed pull requests.
Retenta · Claude Code
An official Retenta skill and MCP connector that teaches multi-run Claude Code workflows to recall state, log progress, and preserve durable findings in a capsule.
Retenta · Codex
An official Retenta skill for giving scheduled and multi-run Codex workflows a durable memory spine through a Retenta MCP capsule.
Cloudflare · Agent Skills clients
Retrieval-first implementation guidance for stateful Cloudflare agents, including Durable Object state, RPC, scheduling, workflows, streaming chat, MCP, and operational gotchas.
Cloudflare · Agent Skills clients
A decision guide for transactional sending, inbound Email Routing, Agents SDK email handlers, REST delivery, domain setup, and deliverability safeguards on Cloudflare.
Cloudflare · Agent Skills clients
Focused patterns for strongly consistent, per-entity coordination with Durable Objects, covering RPC, SQLite storage, alarms, deterministic routing, migrations, and anti-patterns.
Cloudflare · Agent Skills clients
An evidence-led workflow for browser traces, Core Web Vitals, network waterfalls, accessibility snapshots, and codebase-level web performance findings.
Cloudflare · Agent Skills clients
A production profiling workflow for locating CPU and heap hotspots in deployed Workers and Durable Objects and tying pprof evidence back to the smallest code change.
Cloudflare · Agent Skills clients
Operational guidance for project-local Wrangler configuration, development, previews, resource bindings, secrets, permissions, deployment, rollback, and validation.
OpenAI · Codex
An official-source workflow for current OpenAI API guidance, Codex self-knowledge, model selection, and bounded fallback when documentation does not establish a claim.
OpenAI · Codex
A template-first workflow for reproducible experiment and tutorial notebooks, with safe scaffolding, focused cells, validation, and explicit output conventions.
OpenAI · Codex
A render-first process for reading, creating, and reviewing PDFs with layout-aware visual checks and programmatic generation or extraction tools.
OpenAI · Codex
A terminal-first browser automation loop using Playwright CLI snapshots, stable element references, traces, tabs, and artifacts without defaulting to a test-suite rewrite.
OpenAI · Codex
A repository-grounded AppSec workflow for trust boundaries, assets, attacker capabilities, abuse paths, prioritized risk, and evidence-linked mitigations.
OpenAI · Codex
A repeatable CLI workflow for OpenAI speech transcription, optional speaker diarization, known-speaker hints, chunking, and organized output validation.
Anthropic · Claude Code
A two-stage creative workflow that turns an original computational philosophy into reproducible, seeded, interactive p5.js generative art built from a prescribed viewer template.
Anthropic · Claude Code
Opinionated design-lead guidance for subject-specific visual direction, typography, layout, copy, motion restraint, accessibility, and self-critique before frontend implementation.
Anthropic · Claude Code
A four-phase guide to researching, implementing, testing, and evaluating MCP servers with discoverable tools, typed schemas, focused responses, and realistic read-only evals.
Anthropic · Claude Code
A lightweight router for company updates, 3P reports, newsletters, FAQs, incident reports, and leadership communication using format-specific bundled examples.
Anthropic · Claude Code
A detailed production workflow for reading, creating, editing, validating, and visually inspecting PPTX and POTX files through PptxGenJS or direct OOXML operations.
Anthropic · Claude Code
A formula-safe spreadsheet workflow covering XLSX, XLSM, CSV, and TSV creation or editing, recalculation, compatibility constraints, source documentation, and financial-model conventions.
Vercel · Agent Skills clients
Prioritized React API-design patterns for replacing boolean-prop sprawl with compound components, explicit variants, lifted state, generic provider interfaces, and composition.
Vercel · Agent Skills clients
A state-aware deployment decision tree that prefers previews, inspects Git and project links, handles team scope, chooses push or CLI, and documents no-auth fallbacks.
Vercel · Agent Skills clients
A seventy-rule, impact-ranked React and Next.js performance index spanning async waterfalls, bundles, server work, client fetching, rerenders, rendering, and JavaScript hot paths.
Vercel · Agent Skills clients
Prioritized React Native and Expo guidance for list virtualization, Reanimated, native navigation, images, gestures, state subscriptions, monorepos, fonts, and platform UI.
Vercel · Agent Skills clients
A freshness-first UI audit skill that retrieves Vercel's current web-interface rules, checks specified files, and reports concise file-and-line findings.
Cloudflare · Claude Code (plugin marketplace) · OpenAI Codex (plugin marketplace) · Cursor · VS Code / GitHub Copilot · OpenCode · Pi · any Agent Skills client via npx skills
A product-selection router: it maps a stated need (uploads, background jobs, document search, agents, etc.) to Cloudflare products via a need-to-product table and bundled per-product references (AI Gateway, AI Search, Analytics Engine and many more), then tells the agent to load the specific skill or docs. It recommends Workers + Static Assets for all new sites, and if the project uses the beta `cf` CLI (`cloudflare.config.ts`) it switches the agent away from Wrangler guidance to the `cf` docs.
Cloudflare · Claude Code (plugin marketplace) · OpenAI Codex (plugin marketplace) · Cursor · VS Code / GitHub Copilot · OpenCode · Pi · any Agent Skills client via npx skills
Instructs the agent to prefer retrieved docs over training data when writing or reviewing Workers, to keep compatibility dates current, and to enable Workers Logs and Traces. It lists concrete anti-patterns to flag in review (buffering unbounded bodies, hardcoded secrets, Math.random for tokens, un-awaited promises outside ctx.waitUntil, module-level request state, hand-written Env types instead of `wrangler types`, non-constant-time secret comparison) with references for configuration, runtime patterns and platform APIs.
Cloudflare · Claude Code (plugin marketplace) · OpenAI Codex (plugin marketplace) · Cursor · VS Code / GitHub Copilot · OpenCode · Pi · any Agent Skills client via npx skills
Tells the agent to use vinext (Cloudflare's Vite-based reimplementation of the Next.js API) rather than OpenNext for new Next.js projects on Workers, while preserving existing OpenNext setups. For setup, migration or deployment it has the agent install and follow the upstream vinext skills (`npx skills add cloudflare/vinext`).
Cloudflare · Claude Code (plugin marketplace) · OpenAI Codex (plugin marketplace) · Cursor · VS Code / GitHub Copilot · OpenCode · Pi · any Agent Skills client via npx skills
Runs a Core Web Vitals / Lighthouse-style performance audit: the agent first discovers which browser/performance MCP tools are available (e.g. Chrome DevTools MCP), collects traces where possible, and retrieves current thresholds from web.dev and Chrome docs rather than quoting from memory. If trace tools are missing it does source and network analysis and states which measurements it could not collect.
Cloudflare · Claude Code (plugin marketplace) · OpenAI Codex (plugin marketplace) · Cursor · VS Code / GitHub Copilot · OpenCode · Pi · any Agent Skills client via npx skills
A scripted wizard that turns 'set up Turnstile' into a working integration: the agent runs bundled shell scripts (which return JSON) to probe API auth, create a widget, then edits the existing frontend (framework references for HTML, Next.js, Astro, Hugo, SvelteKit) and backend to add server-side Siteverify, and finally runs a real validation pass before reporting success.
WordPress (WordPress.org contributors) · Claude Code · OpenAI Codex · Cursor · VS Code / GitHub Copilot · any Agent Skills client via npx skills (repo README)
A procedure for Gutenberg block work: run the repo triage script, locate blocks, then handle block.json metadata, registration, attributes and serialization, supports, dynamic rendering (render.php / render_callback), deprecations and migrations, viewScriptModule vs viewScript, and @wordpress/scripts / create-block build and test workflows, each backed by a reference file.
WordPress (WordPress.org contributors) · Claude Code · OpenAI Codex · Cursor · VS Code / GitHub Copilot · any Agent Skills client via npx skills (repo README)
Guides plugin architecture and hooks, activation/deactivation/uninstall, the Settings API and admin UI, data storage and cron, release packaging, and security fundamentals (nonces, capability checks, sanitization/escaping, $wpdb->prepare). It starts with a triage script to locate plugin entrypoints and asks for target WP/PHP versions and single-site vs multisite.
WordPress (WordPress.org contributors) · Claude Code · OpenAI Codex · Cursor · VS Code / GitHub Copilot · any Agent Skills client via npx skills (repo README)
Guides building and debugging REST endpoints: register_rest_route and WP_REST_Controller classes, schema and argument validation, permission_callback and authentication modes (cookie+nonce, application passwords), response shaping, register_rest_field/register_meta, and exposing CPTs/taxonomies with show_in_rest. Includes a debugging path for 401/403/404 and nonce issues.
WordPress (WordPress.org contributors) · Claude Code · OpenAI Codex · Cursor · VS Code / GitHub Copilot · any Agent Skills client via npx skills (repo README)
Covers block-theme work: theme.json presets/settings/styles, templates and template parts, patterns, style variations, and Site Editor troubleshooting (style hierarchy, user overrides stored in the DB, caching). It locates theme roots with a triage script and asks where the problem shows up (Site Editor, post editor, frontend).
WordPress (WordPress.org contributors) · Claude Code · OpenAI Codex · Cursor · VS Code / GitHub Copilot · any Agent Skills client via npx skills (repo README)
Operational WordPress work through WP-CLI: safe search-replace for domain/protocol migrations, db export/import, plugin/theme/user management, cron, cache and rewrite flushing, multisite targeting (--url, --network) and wp-cli.yml automation. It requires the agent to establish environment (local/staging/production), site root and safety constraints before running write commands, with a dedicated safety reference.
HashiCorp · Claude Code · OpenAI Codex · Cursor · GitHub Copilot · OpenCode · any Agent Skills client via npx skills
Makes the agent generate and review Terraform HCL to HashiCorp's published style guide: provider and version constraints first, data sources before dependents, resources in dependency order, standard file layout and naming, formatting and variable/output conventions.
HashiCorp · Claude Code · OpenAI Codex · Cursor · GitHub Copilot · OpenCode · any Agent Skills client via npx skills
Guides writing and running Terraform's native tests (.tftest.hcl): run blocks, assertions, plan vs apply tests, mock providers and data sources (Terraform 1.7+), testing module outputs, and troubleshooting test syntax. References cover mocks, CI/CD pipelines (GitHub Actions and GitLab CI) and a full example suite for a VPC module.
Docker · Claude Code · OpenAI Codex (agents/openai.yaml bundled) · Cursor · Gemini CLI · GitHub Copilot · OpenCode · Windsurf · Cline · Kiro · Antigravity · any Agent Skills client via npx skills
Rules and patterns for production Dockerfiles: multi-stage builds, layer-cache ordering, .dockerignore hygiene, non-root users and runtime image size. It ships language templates (Go, Node.js, Python) and a verification checklist, and tells the agent to build the image to verify its changes.
Docker · Claude Code · OpenAI Codex (agents/openai.yaml bundled) · Cursor · Gemini CLI · GitHub Copilot · OpenCode · Windsurf · Cline · Kiro · Antigravity · any Agent Skills client via npx skills
Rules for creating, reviewing and debugging compose.yaml: service definitions, health checks with depends_on conditions, volumes and networks, environment variables, and compose.override.yaml for development. Includes example files, a bad-vs-good comparison and a verification checklist.
Pulumi · Claude Code · OpenAI Codex (agents/openai.yaml bundled) · Cursor · Gemini CLI · VS Code / GitHub Copilot · any Agent Skills client via npx skills
A set of numbered practices for Pulumi TypeScript/Python programs: never create resources inside apply(), pass Outputs directly as inputs, group related resources into ComponentResources with parent: this, encrypt secrets from day one (config --secret / ESC), use aliases when refactoring to avoid replacement, and always preview before deploying (including in CI). Each has incorrect/correct code examples.
Grafana Labs · Claude Code · OpenAI Codex · Cursor · any Agent Skills client via npx skills
Guides instrumenting an app with OpenTelemetry and shipping metrics, logs, traces and profiles to Grafana Cloud or self-hosted Mimir/Loki/Tempo/Pyroscope: SDK auto-instrumentation per language (Go, Java agent, Python, Node.js, .NET), Beyla eBPF for zero-code, Alloy collector configuration, and the Grafana Cloud OTLP gateway with Basic auth.
Google · Claude Code · OpenAI Codex · Antigravity · any Agent Skills client via npx skills
Makes the agent generate or update Kubernetes YAML for GKE Autopilot and Standard under strict rules: explicit namespaces, security contexts, CPU/memory requests and limits, readiness/liveness/startup probes, secret and volume mounts, Gateway API routes, network policies and Spot VM targeting. References cover basic workloads, AI inference, Gateway API and network policy.
Anthropic · Claude Code (plugin marketplace) · Claude.ai · Claude API (Skills API)
A four-phase process for building MCP servers in Python (FastMCP) or TypeScript (MCP SDK): research the target API and MCP best practices, implement tools with good naming, schemas and error messages, review and test, then write an evaluation of realistic questions. Reference files cover MCP best practices, per-language server guides and how to build evaluations.
Anthropic · Claude Code (plugin marketplace) · Claude.ai · Claude API (Skills API)
Guides the full loop of authoring an Agent Skill: draft it, write test prompts, run Claude with and without the skill, grade outputs with grader/comparator/analyzer sub-agent prompts, show results in a generated HTML review viewer, and iterate. It also runs trigger evals that test and rewrite the skill's description for triggering accuracy, then packages the skill.
Anthropic · Claude Code (plugin marketplace) · Claude.ai · Claude API (Skills API)
A reference skill for building on the Claude API and Anthropic SDKs: it first checks the project for other-provider markers and stops if found, detects the language, then routes to per-language docs (Python, TypeScript, Java, Go, C#, curl and more) for streaming, tool use, files, batches, caching and Managed Agents. It carries a dated table of current model IDs and migration rules and warns the agent its training prior on the API is likely stale.
Google (Gemini team) · Claude Code · OpenAI Codex · Cursor · Antigravity · any Agent Skills client via npx skills
Overrides the agent's training data with a current Gemini model list and critical rules, then guides SDK usage for text generation, chat, multimodal input, image/video/speech generation, streaming, function calling, structured output and long-running research tasks. A reference covers migrating from the older generateContent API.
Hugging Face · Claude Code · OpenAI Codex · Cursor · Gemini CLI
Guides fine-tuning language and vision models with TRL (SFT, DPO, GRPO, reward modeling) or Unsloth on Hugging Face Jobs cloud GPUs: dataset validation, hardware and cost estimation, UV scripts in PEP 723 format, Trackio monitoring, saving to the Hub, and GGUF conversion for local inference. Bundled example training scripts are submitted as jobs.
Datadog · Claude Code · OpenAI Codex · Cursor · Gemini CLI · OpenCode · Windsurf · any Agent Skills client via npx skills
Bootstraps LLM evaluators from production traces in Datadog LLM Observability: it samples traces, proposes online LLM-judge evaluators, and after explicit confirmation creates them in Datadog as disabled drafts (never auto-enabled); on request it emits Python SDK code or a framework-agnostic JSON spec instead. It auto-selects between the Datadog MCP server and the `pup` CLI as the backend.
Anthropic · Claude Code (plugin marketplace) · Claude.ai · Claude API (Skills API)
Prompts the agent to act as a design lead with a distinct point of view: identify the subject, audience and primary job of the UI first, then make deliberate, subject-specific choices of palette, typography, layout and hero treatment instead of templated defaults (explicitly avoiding stock patterns such as the big-number-plus-gradient hero).
Anthropic · Claude Code (plugin marketplace) · Claude.ai · Claude API (Skills API)
Has the agent test local web apps by writing native Python Playwright scripts, following a decision tree (static HTML vs dynamic app, server already running or not). It uses a bundled helper to start one or more dev servers, wait for their ports, run the test script and clean up, and examples for console-log capture and element discovery.
Vercel · any Agent Skills client via npx skills (README) · Claude.ai (README)
A prioritized rulebook of about 70 React and Next.js performance rules in 8 categories (async waterfalls, bundle size, server components and caching, re-renders, rendering, effects, etc.), each in its own rules/ file with incorrect vs correct code. The agent applies them when writing, reviewing or refactoring React/Next.js code.
Vercel · any Agent Skills client via npx skills (README) · Claude.ai (README)
Reviews UI files against Vercel's Web Interface Guidelines: the agent fetches the latest guidelines at run time from a raw GitHub URL, reads the files or glob the user names, checks every rule, and reports findings in terse file:line format.
GitHub awesome-copilot (contributed by AmirhosseinHonardoust) · GitHub Copilot (VS Code / Copilot CLI)
A security reviewer for .github/workflows files that reasons about the Actions threat model: script injection from untrusted ${{ }} inputs, privileged triggers (pull_request_target, workflow_run, issue_comment) running fork code, mutable action references vs SHA pinning, over-scoped GITHUB_TOKEN permissions, OIDC and secret handling. Findings follow a bundled report format.
Sentry · Claude Code · Cursor · GitHub Copilot · Cline · any Agent Skills client via npx skills
An exploitation-focused GitHub Actions review: every finding must come with a concrete attack scenario or it is not reported. References encode real attack patterns (pwn requests, expression injection, comment-triggered commands, credential escalation, AI prompt injection via CI, runner infrastructure, supply chain) drawn from the 2025 'HackerBot Claw' campaign analysis.
OpenAI · OpenAI Codex ($skill-installer)
Produces a repository-grounded AppSec threat model in Markdown: the agent extracts components, data stores and integrations from the code, enumerates trust boundaries, assets, attacker capabilities and abuse paths, and proposes mitigations, anchoring every claim to repo evidence and keeping assumptions explicit. It only triggers on an explicit threat-modeling request and follows a bundled output template.
Trail of Bits · Claude Code (plugin marketplace) · OpenAI Codex (agents/openai.yaml bundled) · Antigravity
Runs a Semgrep security scan end to end: detects languages, selects rulesets, presents the plan for explicit user approval, then executes every approved ruleset through a bundled script (metrics off, batched processes) and merges results into one SARIF file. Supports 'run all' and 'important only' (medium-to-high confidence/impact) modes and uses Semgrep Pro cross-file taint analysis when available.
Supabase · Claude Code · Cursor · GitHub Copilot · Cline · any Agent Skills client via npx skills
A Postgres rulebook maintained by Supabase that the agent loads before changing anything in a database: rules across 8 impact-ranked categories (query performance and indexing, connection management and pooling, schema design and column types, RLS policies and their tests, migrations, batch inserts and N+1, JSONB and full-text search). Each rule has incorrect vs correct SQL and query-plan reasoning.
Neon · Claude Code · Cursor · Pi · any Agent Skills client via npx skills
Guides work against an existing Neon Postgres database (branded 'Lakebase Postgres on Neon' in the skill): selecting org/project, pooled vs direct connection strings and driver choice, schema migrations, branching, autoscaling and scale-to-zero, instant restore, read replicas, IP allow lists, logical replication, and Lakebase Search (vector, full-text/BM25, hybrid). It defers new backends, Auth, files and Functions to the parent `neon` skill.
Anthropic · Claude Code (plugin marketplace) · Claude.ai · Claude API (Skills API)
Runs a three-stage co-authoring workflow for specs, proposals, PRDs, RFCs and decision docs: context gathering (the agent asks targeted questions), refinement and structure (iterative drafting section by section), and reader testing (checking the document works for someone without the author's context). The agent offers the workflow when the user starts a substantial writing task.
Sentry · Claude Code · Cursor · GitHub Copilot · Cline · any Agent Skills client via npx skills
Writes or refreshes pull-request titles and bodies as a cover note for reviewers rather than a changelog or file-by-file list: the agent inspects the branch, working tree, existing PR, base branch, commits and full diff with git and gh, then drafts a concise title and body and updates the PR.
No reviewed records match those filters. Clear a filter or browse the trust method to see why thin imports are not published as landing pages.
Skills + governed memory
Keep state between runs with provenance, version history, and a portable knowledge asset.