loadout-ai 0.1.0

package/docs/CATALOG.md

@@ -0,0 +1,78 @@
+# Catalog and upstream credits
+
+Loadout's bundled catalog indexes 50 public upstream repositories. This page gives every project a direct source link and exposes the exact reviewed revision used by Loadout.
+
+Thank you to every maintainer and contributor whose work appears here. Loadout does not own, endorse for its authors, or relicense these projects. Each project remains governed by its upstream license and current terms.
+
+## How to read this page
+
+- **Source** links to the upstream repository.
+- **ID** is Loadout's stable package identifier.
+- **Components** are the component types evidenced in the reviewed repository snapshot; they do not imply every supported agent can install every type.
+- **License** is the SPDX identifier recorded in `catalog/packages.json` from GitHub metadata. **Review required** corresponds to `NOASSERTION`: no SPDX identifier was reported, so users must inspect the upstream terms instead of assuming permission.
+- **Reviewed revision** links to the immutable commit inspected for catalog admission. It is not necessarily the newest upstream commit.
+
+The machine-readable catalog remains the source of truth. Run `loadout catalog --coverage --json` to inspect coverage and `loadout capabilities` to inspect what each local agent adapter can actually manage.
+
+## All 50 credited repositories
+
+| Upstream source                                                                   | Loadout ID                       | Category                 | Reviewed components        | License status      | Reviewed revision                                                                                                       |
+| --------------------------------------------------------------------------------- | -------------------------------- | ------------------------ | -------------------------- | ------------------- | ----------------------------------------------------------------------------------------------------------------------- |
+| [Superpowers](https://github.com/obra/superpowers)                                | `superpowers`                    | engineering-workflow     | `skill`, `plugin`          | `MIT`               | [`d884ae04edeb`](https://github.com/obra/superpowers/tree/d884ae04edebef577e82ff7c4e143debd0bbec99)                     |
+| [Context7](https://github.com/upstash/context7)                                   | `context7`                       | documentation            | `skill`, `mcp`, `plugin`   | `MIT`               | [`eb7ac502f6e8`](https://github.com/upstash/context7/tree/eb7ac502f6e83692363375d3088d550afa298a60)                     |
+| [Playwright MCP](https://github.com/microsoft/playwright-mcp)                     | `playwright-mcp`                 | browser-verification     | `mcp`                      | `Apache-2.0`        | [`5f8fc00210b2`](https://github.com/microsoft/playwright-mcp/tree/5f8fc00210b27b4407c375b59cda4838045d429c)             |
+| [UI UX Pro Max](https://github.com/nextlevelbuilder/ui-ux-pro-max-skill)          | `ui-ux-pro-max`                  | frontend-design          | `skill`, `plugin`          | `MIT`               | [`fe3f65e5d020`](https://github.com/nextlevelbuilder/ui-ux-pro-max-skill/tree/fe3f65e5d0203341b40b7db80650501cd9de4961) |
+| [GitHub MCP Server](https://github.com/github/github-mcp-server)                  | `github-mcp-server`              | source-control           | `mcp`                      | `MIT`               | [`c36e4e4493c7`](https://github.com/github/github-mcp-server/tree/c36e4e4493c76eafe2c1a9cbc5272e8eede29189)             |
+| [OpenAI Skills Catalog](https://github.com/openai/skills)                         | `openai-skills`                  | agent-skills             | `skill`                    | **Review required** | [`49f948faa925`](https://github.com/openai/skills/tree/49f948faa9258a0c61caceaf225e179651397431)                        |
+| [Anthropic Skills](https://github.com/anthropics/skills)                          | `anthropic-skills`               | agent-skills             | `skill`                    | **Review required** | [`9d2f1ae18723`](https://github.com/anthropics/skills/tree/9d2f1ae187231d8199c64b5b762e1bdf2244733d)                    |
+| [Agent Skills Marketplace](https://github.com/wshobson/agents)                    | `wshobson-agents`                | agent-skills             | `skill`, `plugin`          | `MIT`               | [`b6af37110581`](https://github.com/wshobson/agents/tree/b6af3711058190e4b5c5274b9758498fe626ec5a)                      |
+| [Vercel Agent Skills](https://github.com/vercel-labs/agent-skills)                | `vercel-agent-skills`            | frontend-development     | `skill`                    | **Review required** | [`f8a72b960372`](https://github.com/vercel-labs/agent-skills/tree/f8a72b9603728bb92a217a879b7e62e43ad76c81)             |
+| [Vercel Skills](https://github.com/vercel-labs/skills)                            | `vercel-skills`                  | skill-discovery          | `skill`                    | **Review required** | [`5527c09adc36`](https://github.com/vercel-labs/skills/tree/5527c09adc367612b0bffd9c80e3bc28a6b01b6d)                   |
+| [Cloudflare MCP Server](https://github.com/cloudflare/mcp-server-cloudflare)      | `cloudflare-mcp-server`          | cloud-platform           | `mcp`                      | `Apache-2.0`        | [`52c633e37684`](https://github.com/cloudflare/mcp-server-cloudflare/tree/52c633e37684fadb94ae236f74909b9bbefc0db8)     |
+| [Supabase MCP](https://github.com/supabase/mcp)                                   | `supabase-mcp`                   | database                 | `mcp`                      | `Apache-2.0`        | [`9f0396a1b367`](https://github.com/supabase/mcp/tree/9f0396a1b367dfa6a424f4686a21701c8f05b1c3)                         |
+| [Sentry MCP](https://github.com/getsentry/sentry-mcp)                             | `sentry-mcp`                     | observability            | `skill`, `mcp`             | **Review required** | [`ce099fd25197`](https://github.com/getsentry/sentry-mcp/tree/ce099fd251973774116749016847b78a6049ba0a)                 |
+| [Exa MCP Server](https://github.com/exa-labs/exa-mcp-server)                      | `exa-mcp-server`                 | web-research             | `skill`, `mcp`, `plugin`   | `MIT`               | [`8823cbea80b6`](https://github.com/exa-labs/exa-mcp-server/tree/8823cbea80b6ec5999c6217cccd91660fe953b97)              |
+| [Firecrawl MCP Server](https://github.com/firecrawl/firecrawl-mcp-server)         | `firecrawl-mcp-server`           | web-research             | `mcp`                      | `MIT`               | [`3eb1115b1f28`](https://github.com/firecrawl/firecrawl-mcp-server/tree/3eb1115b1f2883ff2fb74e61b5c4acf5a9ac0fb0)       |
+| [Azure DevOps MCP](https://github.com/microsoft/azure-devops-mcp)                 | `azure-devops-mcp`               | devops                   | `mcp`                      | `MIT`               | [`0a08dc649c95`](https://github.com/microsoft/azure-devops-mcp/tree/0a08dc649c951208eda942154c66187dec9985bd)           |
+| [Docker MCP Gateway](https://github.com/docker/mcp-gateway)                       | `docker-mcp-gateway`             | containers               | `skill`, `mcp`             | `MIT`               | [`6ce7ff38e75e`](https://github.com/docker/mcp-gateway/tree/6ce7ff38e75e1c62ccefd9882145c4a2ab3134ed)                   |
+| [Hugging Face MCP Server](https://github.com/huggingface/hf-mcp-server)           | `huggingface-mcp-server`         | machine-learning         | `mcp`                      | `MIT`               | [`2ec6f0723591`](https://github.com/huggingface/hf-mcp-server/tree/2ec6f0723591214ac647ceb0a31c34221b9a317e)            |
+| [Awesome Copilot](https://github.com/github/awesome-copilot)                      | `awesome-copilot`                | agent-skills             | `skill`, `agent`, `mcp`    | `MIT`               | [`2c2461a7fa38`](https://github.com/github/awesome-copilot/tree/2c2461a7fa383f664bb75546f03a2c6087f3819d)               |
+| [OpenAI Codex Skills](https://github.com/openai/codex)                            | `openai-codex-skills`            | agent-skills             | `skill`, `agent`           | `Apache-2.0`        | [`2e1607ee2fa8`](https://github.com/openai/codex/tree/2e1607ee2fa8099a233df7437adee5f16a741905)                         |
+| [Ponytail](https://github.com/DietrichGebert/ponytail)                            | `ponytail`                       | code-simplification      | `skill`, `plugin`          | `MIT`               | [`16f29800fd26`](https://github.com/DietrichGebert/ponytail/tree/16f29800fd2681bdf24f3eb4ccffe38be3baec6b)              |
+| [Addy Osmani Agent Skills](https://github.com/addyosmani/agent-skills)            | `addyosmani-agent-skills`        | software-engineering     | `skill`, `plugin`          | `MIT`               | [`c1974de476a3`](https://github.com/addyosmani/agent-skills/tree/c1974de476a39cb002a3b8e51e6a7e8e57b808c6)              |
+| [Scientific Agent Skills](https://github.com/K-Dense-AI/scientific-agent-skills)  | `scientific-agent-skills`        | scientific-research      | `skill`                    | `MIT`               | [`3f825caafe14`](https://github.com/K-Dense-AI/scientific-agent-skills/tree/3f825caafe149b7853ec8c4d1dd7f4553ea6b2a5)   |
+| [Planning with Files](https://github.com/OthmanAdi/planning-with-files)           | `planning-with-files`            | planning                 | `skill`, `plugin`          | `MIT`               | [`f90780c92f05`](https://github.com/OthmanAdi/planning-with-files/tree/f90780c92f0506d21c5f6c4865ce95517a8b1964)        |
+| [PM Skills](https://github.com/phuryn/pm-skills)                                  | `pm-skills`                      | product-management       | `skill`, `plugin`          | `MIT`               | [`18468a95b427`](https://github.com/phuryn/pm-skills/tree/18468a95b427e70e258b51389796367c6f684e7d)                     |
+| [Baoyu Skills](https://github.com/JimLiu/baoyu-skills)                            | `baoyu-skills`                   | content-creation         | `skill`                    | `MIT`               | [`6b7a2e417500`](https://github.com/JimLiu/baoyu-skills/tree/6b7a2e417500561a5ecdd0b168332f4142584617)                  |
+| [Trail of Bits Skills](https://github.com/trailofbits/skills)                     | `trail-of-bits-skills`           | security                 | `skill`, `plugin`          | `CC-BY-SA-4.0`      | [`cfe5d7b1619e`](https://github.com/trailofbits/skills/tree/cfe5d7b1619e47fb5b38b7e2561dad7e5f1e89af)                   |
+| [Antfu Skills](https://github.com/antfu/skills)                                   | `antfu-skills`                   | frontend-development     | `skill`                    | `MIT`               | [`a74f281a27da`](https://github.com/antfu/skills/tree/a74f281a27dadc02397bc1a174b0f2c97531b6ae)                         |
+| [.NET Skills](https://github.com/dotnet/skills)                                   | `dotnet-skills`                  | dotnet-development       | `skill`, `agent`, `plugin` | `MIT`               | [`79a2ada302fe`](https://github.com/dotnet/skills/tree/79a2ada302fe8feb884ab33e1bfc2bddae3ee7ce)                        |
+| [Microsoft Skills](https://github.com/microsoft/skills)                           | `microsoft-skills`               | cloud-development        | `skill`, `agent`, `plugin` | `MIT`               | [`e19efc27887e`](https://github.com/microsoft/skills/tree/e19efc27887e5f1a6380b9135cdbdbf10648fba7)                     |
+| [Web Quality Skills](https://github.com/addyosmani/web-quality-skills)            | `web-quality-skills`             | web-quality              | `skill`, `plugin`          | `MIT`               | [`95d6e255afe1`](https://github.com/addyosmani/web-quality-skills/tree/95d6e255afe1596b557d7a8498517884438f5b3a)        |
+| [Softaworks Agent Toolkit](https://github.com/softaworks/agent-toolkit)           | `softaworks-agent-toolkit`       | engineering-productivity | `skill`                    | `MIT`               | [`3027f20f3181`](https://github.com/softaworks/agent-toolkit/tree/3027f20f3181758385a1bb8c022d4041dfb4de84)             |
+| [Draw.io Skill](https://github.com/Agents365-ai/drawio-skill)                     | `drawio-skill`                   | diagramming              | `skill`                    | `MIT`               | [`9f5292577033`](https://github.com/Agents365-ai/drawio-skill/tree/9f52925770334e9bd2794e40a7fa64c234e9f545)            |
+| [Chrome DevTools MCP](https://github.com/ChromeDevTools/chrome-devtools-mcp)      | `chrome-devtools-mcp`            | browser-debugging        | `skill`, `mcp`, `plugin`   | `Apache-2.0`        | [`9a7e04daa2cd`](https://github.com/ChromeDevTools/chrome-devtools-mcp/tree/9a7e04daa2cd031afe469d4503d6baf58b9b2cea)   |
+| [Serena](https://github.com/oraios/serena)                                        | `serena`                         | code-intelligence        | `mcp`                      | `MIT`               | [`465c5eb62788`](https://github.com/oraios/serena/tree/465c5eb627883ec310c5b7e643fe445af84589b3)                        |
+| [Model Context Protocol Servers](https://github.com/modelcontextprotocol/servers) | `model-context-protocol-servers` | mcp-reference            | `mcp`                      | **Review required** | [`d31124c98240`](https://github.com/modelcontextprotocol/servers/tree/d31124c982401739917fd817c2a59db344529c16)         |
+| [AWS MCP Servers](https://github.com/awslabs/mcp)                                 | `aws-mcp-servers`                | cloud-platform           | `mcp`                      | `Apache-2.0`        | [`778b5a5db8a6`](https://github.com/awslabs/mcp/tree/778b5a5db8a6f5b92ddb355e59b398544dbf6630)                          |
+| [DBHub](https://github.com/bytebase/dbhub)                                        | `dbhub`                          | database                 | `skill`, `mcp`             | `MIT`               | [`ec8efede6b8b`](https://github.com/bytebase/dbhub/tree/ec8efede6b8b47dabceb822022954852fc420490)                       |
+| [FastAPI MCP](https://github.com/tadata-org/fastapi_mcp)                          | `fastapi-mcp`                    | api-development          | `mcp`                      | `MIT`               | [`e5cad13cabfc`](https://github.com/tadata-org/fastapi_mcp/tree/e5cad13cabfc725bbcb047e526816d887d96da62)               |
+| [Browser MCP](https://github.com/BrowserMCP/mcp)                                  | `browser-mcp`                    | browser-automation       | `mcp`                      | `Apache-2.0`        | [`9db12f2b4f61`](https://github.com/BrowserMCP/mcp/tree/9db12f2b4f61294f0bc11708986abc47db539d6c)                       |
+| [AntV Chart MCP](https://github.com/antvis/mcp-server-chart)                      | `antvis-chart-mcp`               | data-visualization       | `mcp`                      | `MIT`               | [`33178096abfa`](https://github.com/antvis/mcp-server-chart/tree/33178096abfa58f245b456f5bacb18bc37f5cbaf)              |
+| [Excel MCP Server](https://github.com/haris-musa/excel-mcp-server)                | `excel-mcp-server`               | spreadsheets             | `mcp`                      | `MIT`               | [`f51340ecd577`](https://github.com/haris-musa/excel-mcp-server/tree/f51340ecd5778952405044b203d3a2d4c8a46833)          |
+| [arXiv MCP Server](https://github.com/blazickjp/arxiv-mcp-server)                 | `arxiv-mcp-server`               | academic-research        | `skill`, `mcp`, `plugin`   | `Apache-2.0`        | [`d58901760d7e`](https://github.com/blazickjp/arxiv-mcp-server/tree/d58901760d7ede4adb162eaba1725209a933f100)           |
+| [Google Workspace MCP](https://github.com/taylorwilsdon/google_workspace_mcp)     | `google-workspace-mcp`           | productivity             | `skill`, `mcp`, `plugin`   | `MIT`               | [`3f19c7d4a265`](https://github.com/taylorwilsdon/google_workspace_mcp/tree/3f19c7d4a2654238e5fed1b767174fac5615555b)   |
+| [MongoDB MCP Server](https://github.com/mongodb-js/mongodb-mcp-server)            | `mongodb-mcp-server`             | database                 | `mcp`                      | `Apache-2.0`        | [`f45d7c777d88`](https://github.com/mongodb-js/mongodb-mcp-server/tree/f45d7c777d88d77069da36b3133e68901dc0697a)        |
+| [Redis MCP Server](https://github.com/redis/mcp-redis)                            | `redis-mcp-server`               | database                 | `mcp`                      | `MIT`               | [`53df2bcb7754`](https://github.com/redis/mcp-redis/tree/53df2bcb7754a393f53b6c880ab710b95fefa343)                      |
+| [Stripe AI](https://github.com/stripe/ai)                                         | `stripe-ai`                      | payments                 | `skill`, `mcp`, `plugin`   | `MIT`               | [`e238b951665c`](https://github.com/stripe/ai/tree/e238b951665cba5260d70aa80e5021def27a61d3)                            |
+| [MCP Toolbox for Databases](https://github.com/googleapis/mcp-toolbox)            | `mcp-toolbox-for-databases`      | database                 | `mcp`                      | `Apache-2.0`        | [`b574b071fc27`](https://github.com/googleapis/mcp-toolbox/tree/b574b071fc27f717e90b4362864b1086be0c8995)               |
+| [Browserbase MCP Server](https://github.com/browserbase/mcp-server-browserbase)   | `browserbase-mcp-server`         | browser-automation       | `mcp`                      | `Apache-2.0`        | [`1e196b3d3c4d`](https://github.com/browserbase/mcp-server-browserbase/tree/1e196b3d3c4dc70944e0d19038dd9eb3608b207a)   |
+| [Bright Data MCP](https://github.com/brightdata/brightdata-mcp)                   | `bright-data-mcp`                | web-research             | `mcp`                      | `MIT`               | [`adbfdd49b880`](https://github.com/brightdata/brightdata-mcp/tree/adbfdd49b8805136662a264490118c2688da3e29)            |
+
+## Catalog maintenance
+
+`loadout catalog --refresh` refreshes mutable observations such as stars, topics, releases, and release-asset downloads in local state. It does not silently rewrite the reviewed commit or component evidence in the bundled catalog. Re-running it on the same UTC date replaces that day's observation rather than inventing velocity.
+
+A catalog revision should be reviewed again before changing the pin. New projects found by daily discovery remain candidates until they pass provenance, component, overlap, license, platform, and safety review.
+
+For policy details, see [Catalog ranking and conflict policy](./CATALOG_POLICY.md), [Community discovery](./COMMUNITY_DISCOVERY.md), and [Provenance and comparison](./PROVENANCE_AND_COMPARISON.md).

package/docs/CATALOG_POLICY.md

@@ -0,0 +1,41 @@
+# Catalog ranking and conflict policy
+
+Loadout's ranking is an explanation, not a claim that one agent extension is
+universally "best." It compares candidates only when they serve the same
+capability family. An impressive star count cannot promote an uninspected
+package over a safety or provenance failure.
+
+Trust and ranking are deliberately separate. Every package is labelled as
+`discovered`, `inspected`, `human-reviewed`, `benchmarked`, or `recommended`.
+The bundled 50-repository release is technically screened and pinned, but that
+does not mean all 50 are universal winners or have received human review. The
+current Stable recommendation is narrower: 30 selected skills from four
+immutable, SPDX-identified sources whose prepared plan needs no additional
+static-risk approval.
+
+The current score has five bounded factors: logarithmic GitHub stars (30),
+measured momentum (20), code-push maintenance (20), evidenced components and
+platforms (15), and immutable source/license/review evidence (15). A factor
+with no stored evidence receives zero points. In particular, one star snapshot
+does not establish momentum, `updated_at` metadata is not code maintenance,
+and README claims/topics are not independent trust evidence.
+
+This limits obvious gaming and bias: stars are capped and logarithmic; missing
+data does not gain a neutral score; archived packages are not auto-selected;
+official publisher status is not a security guarantee; and deterministic ID
+ordering breaks exact ties. The score must not be used to compare unrelated
+categories, infer security, or make a personal recommendation from protected
+or private data.
+
+Conflict families are maintained policy, separate from filesystem collision
+checks. The current real-catalog families are broad skill collections and web
+research MCP servers. They are soft overlaps: Stable chooses a deliberately
+curated skill subset, while Power, Maximum, and Custom retain broader choices
+with a plain-language warning. Power filters collections at skill granularity;
+Maximum stores the full technically screened set disabled in the library. A hard
+family is only added with evidence of real incompatibility; it blocks every
+profile until a person removes all but one candidate.
+
+Run `loadout setup --mode stable` for the recommended preview or `loadout plan
+--mode power` for the broader set. The resolver runs before changing agent
+directories.

package/docs/COMMUNITY_DISCOVERY.md

@@ -0,0 +1,14 @@
+# Community discovery boundary
+
+Loadout's Hacker News connector calls the documented public Firebase API only:
+`/v0/topstories.json` followed by `/v0/item/<id>.json`. It does not scrape Hacker
+News HTML, sign in, post, vote, retain author profiles, or poll in the background.
+
+The connector reports only a GitHub `owner/repository` mentioned by a current story,
+the story score and comment count, and direct links to the story and discussion. A
+community mention is an early-discovery signal, not a trust signal: it never changes
+the verified catalog, starts a clone, or makes an installation eligible. The user must
+still use `plan`, inspect the static safety report, and explicitly approve an install.
+
+This design makes a viral repository visible quickly without turning social momentum
+into an automatic recommendation or a supply-chain execution path.

package/docs/COMPATIBILITY_POLICY.md

@@ -0,0 +1,22 @@
+# Compatibility policy
+
+Loadout uses the same adapter matrix in inspection and installation. Each
+agent/component cell is one of:
+
+- **native** — Loadout writes the documented component format to the
+  agent-owned location.
+- **adapted** — Loadout has a reviewed, tested mapping to an agent-specific
+  layout or explicit config writer. The plan warns about the adaptation.
+- **unsupported** — Loadout does not infer a location, convert the component,
+  or write anything.
+
+All supported platforms (Windows, macOS, and Linux) use this policy. WSL is
+Linux-side only: Loadout manages the POSIX home directory and never silently
+rewrites a Windows profile. Agent adapters are deliberately conservative:
+skills are native wherever a documented skill directory exists; commands,
+rules, and agents are native only where the adapter supplies a concrete
+directory; MCP is config-scoped and adapted only through its explicit JSON or
+Codex TOML planner; and plugin manifests are inspected/normalized but their
+runtime hooks are never converted or executed.
+
+Use `loadout capabilities --inspect` to view the matrix and local inventory.

package/docs/CONVERSION_AND_SANDBOX.md

@@ -0,0 +1,27 @@
+# Conversion loss reports and installer sandbox boundary
+
+## Cross-agent conversion
+
+The only safe initial conversion target is a static instruction file. A source skill
+can be copied to an agent's documented native skill directory when that adapter marks
+skills `native`. For a target that supports only commands, Loadout may create one
+explicit command wrapper pointing at the same static instructions.
+
+Every conversion emits a loss report containing source component, target component,
+preserved fields, dropped fields, and the reason. Hooks, subagents, lifecycle scripts,
+dynamic variables, permissions, and MCP servers are never silently converted. A loss
+report with dropped executable behavior requires user approval and defaults to no
+write.
+
+## Third-party installer threat model
+
+An installer can read secrets, write outside its target, alter shell startup files,
+open network connections, launch processes, or exploit the host. Therefore source
+repository scripts are never run on the host. Static package parsing is the normal
+path.
+
+If a future disposable sandbox runner is enabled, it must use a fresh filesystem,
+read-only source mount, no host home mount, no inherited environment variables, no
+Docker socket, restricted network disabled by default, bounded CPU/memory/time, and
+an exported file manifest. Its output is an untrusted proposal that still goes through
+the normal plan/safety/approval path. Failure or timeout discards the sandbox.

package/docs/CREDENTIAL_AND_UPDATE_POLICY.md

@@ -0,0 +1,31 @@
+# Credential and autonomous-update policy
+
+## Credential interface
+
+The core accepts `CredentialReference` values only: an environment-variable name or
+an OS-keychain service/account reference. Raw values cannot appear in config schemas,
+logs, plans, lockfiles, exports, or telemetry. Backend implementations must expose
+`get`, `set`, and `delete` by reference; `get` returns a transient value to the direct
+provider client and is never serializable. Unsupported platforms fail closed.
+
+macOS maps to Keychain Services, Windows to Credential Manager, and Linux to Secret
+Service/libsecret. The user controls creation and deletion; Loadout never creates a
+credential backend from a repository manifest.
+
+## Update policy
+
+The default is `manual`: detect and show diffs only. `canary` permits a selected,
+low-risk package to install into a disposable or explicitly named canary profile,
+then requires a static verification pass before human promotion. `automatic` is not
+available for hooks, executables, plugins, MCP configurations, new domains, new
+environment references, unsigned catalogs, or packages without a prior successful
+canary.
+
+Every promotion stores a snapshot and an append-only local audit event. Any failed
+verification restores the snapshot and quarantines the candidate commit. A user can
+revoke a policy at any time; revocation prevents future mutations but never deletes
+existing snapshots.
+
+The local `loadout canary` command implements the non-mutating static gate. A
+transaction layer must provide verification and promotion callbacks before a
+candidate can be promoted; the command itself never installs a candidate.

package/docs/DISCOVERED.md

@@ -0,0 +1,280 @@
+# Daily repository discovery
+
+_Generated from the GitHub REST API on 2026-07-16. Do not edit by hand._
+
+This is a discovery feed, not a list of recommendations. A repository appearing here is **not installed, trusted, or promoted into Loadout's reviewed catalog automatically**. Catalog admission still requires provenance, license, safety, compatibility, and capability review.
+
+## Today at a glance
+
+- 242 repositories appeared across 8 bounded searches.
+- 219 are uncataloged review candidates; 219 were first observed today.
+- 23 already exist in the reviewed catalog.
+- 0 candidates have a star-velocity measurement based on at least one full day of observations.
+- Historical observations are bounded to 90 per repository and candidates are retained for 180 days.
+
+Candidates are ordered by an observed daily star change when Loadout has observations at least one day apart. Until that exists, ordering falls back to **lifetime stars divided by repository age** (with a one-day floor). The lifetime average helps surface young breakouts on day one, but it is not evidence of current growth.
+
+## Uncataloged candidates
+
+| Repository | Stars | Observed velocity | Lifetime star average | Created | License API value | Matched searches | First seen |
+| --- | ---: | ---: | ---: | --- | --- | --- | --- |
+| [affaan-m/ECC](https://github.com/affaan-m/ECC) | 230,262 | — | 1279.5/day | 2026-01-18 | MIT | ai-agents, claude-code, mcp, recent-agent-skills | 2026-07-16 |
+| [nexu-io/open-design](https://github.com/nexu-io/open-design) | 78,794 | — | 987.2/day | 2026-04-28 | Apache-2.0 | agent-skills, ai-agents | 2026-07-16 |
+| [JuliusBrussee/caveman](https://github.com/JuliusBrussee/caveman) | 90,025 | — | 869.1/day | 2026-04-04 | MIT | claude-code | 2026-07-16 |
+| [Graphify-Labs/graphify](https://github.com/Graphify-Labs/graphify) | 88,218 | — | 845.5/day | 2026-04-03 | MIT | claude-code, codex | 2026-07-16 |
+| [Egonex-AI/Understand-Anything](https://github.com/Egonex-AI/Understand-Anything) | 74,442 | — | 600.8/day | 2026-03-15 | MIT | claude-code, codex | 2026-07-16 |
+| [NousResearch/hermes-agent](https://github.com/NousResearch/hermes-agent) | 215,699 | — | 600.7/day | 2025-07-22 | MIT | ai-agents, claude-code, codex | 2026-07-16 |
+| [santifer/career-ops](https://github.com/santifer/career-ops) | 60,297 | — | 584.1/day | 2026-04-04 | MIT | claude-code | 2026-07-16 |
+| [MemPalace/mempalace](https://github.com/MemPalace/mempalace) | 57,377 | — | 557.3/day | 2026-04-05 | MIT | mcp | 2026-07-16 |
+| [Leonxlnx/taste-skill](https://github.com/Leonxlnx/taste-skill) | 64,039 | — | 435.4/day | 2026-02-19 | MIT | claude-code, codex | 2026-07-16 |
+| [rtk-ai/rtk](https://github.com/rtk-ai/rtk) | 71,332 | — | 406.9/day | 2026-01-22 | Apache-2.0 | claude-code | 2026-07-16 |
+| [Panniantong/Agent-Reach](https://github.com/Panniantong/Agent-Reach) | 57,042 | — | 399.1/day | 2026-02-24 | MIT | claude-code, mcp | 2026-07-16 |
+| [calesthio/OpenMontage](https://github.com/calesthio/OpenMontage) | 39,122 | — | 357.7/day | 2026-03-29 | AGPL-3.0 | recent-agent-skills | 2026-07-16 |
+| [farion1231/cc-switch](https://github.com/farion1231/cc-switch) | 117,785 | — | 340.0/day | 2025-08-04 | MIT | claude-code, codex, mcp | 2026-07-16 |
+| [koala73/worldmonitor](https://github.com/koala73/worldmonitor) | 61,924 | — | 327.2/day | 2026-01-08 | NOASSERTION | mcp | 2026-07-16 |
+| [rohitg00/ai-engineering-from-scratch](https://github.com/rohitg00/ai-engineering-from-scratch) | 38,462 | — | 319.9/day | 2026-03-18 | MIT | ai-agents, mcp | 2026-07-16 |
+| [headroomlabs-ai/headroom](https://github.com/headroomlabs-ai/headroom) | 59,437 | — | 312.6/day | 2026-01-07 | Apache-2.0 | claude-code, mcp | 2026-07-16 |
+| [mvanhorn/last30days-skill](https://github.com/mvanhorn/last30days-skill) | 52,400 | — | 300.9/day | 2026-01-23 | MIT | claude-code, recent-agent-skills | 2026-07-16 |
+| [VoltAgent/awesome-openclaw-skills](https://github.com/VoltAgent/awesome-openclaw-skills) | 51,260 | — | 297.0/day | 2026-01-25 | MIT | agent-skills | 2026-07-16 |
+| [code-yeongyu/oh-my-openagent](https://github.com/code-yeongyu/oh-my-openagent) | 65,944 | — | 291.9/day | 2025-12-03 | NOASSERTION | ai-agents, codex | 2026-07-16 |
+| [HKUDS/nanobot](https://github.com/HKUDS/nanobot) | 45,709 | — | 275.9/day | 2026-02-01 | MIT | ai-agents, claude-code, codex | 2026-07-16 |
+| [heygen-com/hyperframes](https://github.com/heygen-com/hyperframes) | 35,554 | — | 275.8/day | 2026-03-10 | Apache-2.0 | mcp | 2026-07-16 |
+| [thedotmack/claude-mem](https://github.com/thedotmack/claude-mem) | 87,450 | — | 274.0/day | 2025-08-31 | Apache-2.0 | ai-agents, claude-code | 2026-07-16 |
+| [Imbad0202/academic-research-skills](https://github.com/Imbad0202/academic-research-skills) | 38,058 | — | 270.4/day | 2026-02-26 | NOASSERTION | claude-code | 2026-07-16 |
+| [op7418/guizang-ppt-skill](https://github.com/op7418/guizang-ppt-skill) | 21,557 | — | 255.3/day | 2026-04-23 | AGPL-3.0 | codex, recent-agent-skills | 2026-07-16 |
+| [ComposioHQ/awesome-claude-skills](https://github.com/ComposioHQ/awesome-claude-skills) | 67,859 | — | 248.8/day | 2025-10-17 | NOASSERTION | agent-skills, ai-agents, claude-code, codex, mcp | 2026-07-16 |
+| [alchaincyf/huashu-design](https://github.com/alchaincyf/huashu-design) | 21,532 | — | 243.6/day | 2026-04-19 | MIT | recent-agent-skills | 2026-07-16 |
+| [shanraisshan/claude-code-best-practice](https://github.com/shanraisshan/claude-code-best-practice) | 62,801 | — | 243.0/day | 2025-10-31 | MIT | ai-agents, claude-code | 2026-07-16 |
+| [sickn33/agentic-awesome-skills](https://github.com/sickn33/agentic-awesome-skills) | 43,386 | — | 236.7/day | 2026-01-14 | MIT | agent-skills, ai-agents, claude-code, mcp | 2026-07-16 |
+| [google-gemini/gemini-cli](https://github.com/google-gemini/gemini-cli) | 106,018 | — | 232.9/day | 2025-04-17 | Apache-2.0 | ai-agents | 2026-07-16 |
+| [JCodesMore/ai-website-cloner-template](https://github.com/JCodesMore/ai-website-cloner-template) | 28,528 | — | 227.3/day | 2026-03-13 | MIT | ai-agents, claude-code | 2026-07-16 |
+| [DeusData/codebase-memory-mcp](https://github.com/DeusData/codebase-memory-mcp) | 32,033 | — | 225.5/day | 2026-02-24 | MIT | claude-code, codex, mcp, model-context-protocol, recent-mcp | 2026-07-16 |
+| [coreyhaines31/marketingskills](https://github.com/coreyhaines31/marketingskills) | 40,058 | — | 219.9/day | 2026-01-15 | MIT | codex | 2026-07-16 |
+| [googleworkspace/cli](https://github.com/googleworkspace/cli) | 29,743 | — | 218.4/day | 2026-03-02 | Apache-2.0 | agent-skills, recent-agent-skills | 2026-07-16 |
+| [kepano/obsidian-skills](https://github.com/kepano/obsidian-skills) | 42,204 | — | 216.3/day | 2026-01-02 | MIT | codex | 2026-07-16 |
+| [Yeachan-Heo/oh-my-claudecode](https://github.com/Yeachan-Heo/oh-my-claudecode) | 37,810 | — | 200.2/day | 2026-01-09 | MIT | ai-agents, claude-code | 2026-07-16 |
+| [CloakHQ/CloakBrowser](https://github.com/CloakHQ/CloakBrowser) | 28,414 | — | 196.4/day | 2026-02-22 | MIT | ai-agents | 2026-07-16 |
+| [shareAI-lab/learn-claude-code](https://github.com/shareAI-lab/learn-claude-code) | 71,195 | — | 186.2/day | 2025-06-29 | MIT | claude-code | 2026-07-16 |
+| [microsoft/SkillOpt](https://github.com/microsoft/SkillOpt) | 12,884 | — | 184.8/day | 2026-05-08 | MIT | agent-skills, recent-agent-skills | 2026-07-16 |
+| [firecrawl/firecrawl](https://github.com/firecrawl/firecrawl) | 151,734 | — | 184.6/day | 2024-04-15 | AGPL-3.0 | ai-agents | 2026-07-16 |
+| [nanocoai/nanoclaw](https://github.com/nanocoai/nanoclaw) | 30,264 | — | 181.9/day | 2026-01-31 | MIT | ai-agents, claude-code | 2026-07-16 |
+| [mukul975/Anthropic-Cybersecurity-Skills](https://github.com/mukul975/Anthropic-Cybersecurity-Skills) | 25,680 | — | 181.4/day | 2026-02-25 | Apache-2.0 | recent-agent-skills | 2026-07-16 |
+| [rohitg00/agentmemory](https://github.com/rohitg00/agentmemory) | 25,204 | — | 177.9/day | 2026-02-25 | Apache-2.0 | codex | 2026-07-16 |
+| [bytedance/deer-flow](https://github.com/bytedance/deer-flow) | 77,173 | — | 177.1/day | 2025-05-07 | MIT | ai-agents | 2026-07-16 |
+| [alibaba/open-code-review](https://github.com/alibaba/open-code-review) | 10,604 | — | 177.0/day | 2026-05-18 | Apache-2.0 | agent-skills | 2026-07-16 |
+| [browser-use/browser-use](https://github.com/browser-use/browser-use) | 105,018 | — | 168.5/day | 2024-10-31 | MIT | ai-agents | 2026-07-16 |
+| [stablyai/orca](https://github.com/stablyai/orca) | 20,300 | — | 166.6/day | 2026-03-17 | MIT | codex | 2026-07-16 |
+| [luongnv89/claude-howto](https://github.com/luongnv89/claude-howto) | 39,901 | — | 158.9/day | 2025-11-07 | MIT | claude-code | 2026-07-16 |
+| [ruvnet/ruflo](https://github.com/ruvnet/ruflo) | 64,598 | — | 157.9/day | 2025-06-02 | MIT | ai-agents, claude-code, codex | 2026-07-16 |
+| [Wei-Shaw/sub2api](https://github.com/Wei-Shaw/sub2api) | 32,467 | — | 153.9/day | 2025-12-18 | LGPL-3.0 | claude-code, codex | 2026-07-16 |
+| [ogulcancelik/herdr](https://github.com/ogulcancelik/herdr) | 17,020 | — | 153.0/day | 2026-03-27 | NOASSERTION | codex | 2026-07-16 |
+| [punkpeye/awesome-mcp-servers](https://github.com/punkpeye/awesome-mcp-servers) | 90,834 | — | 153.0/day | 2024-11-30 | MIT | mcp | 2026-07-16 |
+| [zarazhangrui/frontend-slides](https://github.com/zarazhangrui/frontend-slides) | 25,724 | — | 151.3/day | 2026-01-28 | MIT | recent-agent-skills | 2026-07-16 |
+| [oomol-lab/open-connector](https://github.com/oomol-lab/open-connector) | 2,646 | — | 150.8/day | 2026-06-29 | Apache-2.0 | model-context-protocol, recent-mcp | 2026-07-16 |
+| [Donchitos/Claude-Code-Game-Studios](https://github.com/Donchitos/Claude-Code-Game-Studios) | 23,044 | — | 148.9/day | 2026-02-12 | MIT | recent-agent-skills | 2026-07-16 |
+| [iOfficeAI/OfficeCLI](https://github.com/iOfficeAI/OfficeCLI) | 18,111 | — | 146.4/day | 2026-03-15 | Apache-2.0 | codex | 2026-07-16 |
+| [manaflow-ai/cmux](https://github.com/manaflow-ai/cmux) | 24,583 | — | 145.0/day | 2026-01-28 | NOASSERTION | codex | 2026-07-16 |
+| [tanweai/pua](https://github.com/tanweai/pua) | 18,823 | — | 144.5/day | 2026-03-08 | NOASSERTION | recent-agent-skills | 2026-07-16 |
+| [open-webui/open-webui](https://github.com/open-webui/open-webui) | 145,608 | — | 143.6/day | 2023-10-06 | NOASSERTION | mcp | 2026-07-16 |
+| [tirth8205/code-review-graph](https://github.com/tirth8205/code-review-graph) | 19,529 | — | 139.2/day | 2026-02-26 | MIT | recent-mcp | 2026-07-16 |
+| [google/skills](https://github.com/google/skills) | 14,864 | — | 138.6/day | 2026-03-31 | Apache-2.0 | recent-agent-skills | 2026-07-16 |
+| [larksuite/cli](https://github.com/larksuite/cli) | 15,668 | — | 137.8/day | 2026-03-25 | MIT | recent-agent-skills | 2026-07-16 |
+| [sansan0/TrendRadar](https://github.com/sansan0/TrendRadar) | 60,593 | — | 136.3/day | 2025-04-28 | GPL-3.0 | mcp | 2026-07-16 |
+| [anthropics/claude-plugins-official](https://github.com/anthropics/claude-plugins-official) | 32,196 | — | 135.1/day | 2025-11-20 | Apache-2.0 | claude-code, mcp | 2026-07-16 |
+| [asgeirtj/system_prompts_leaks](https://github.com/asgeirtj/system_prompts_leaks) | 58,243 | — | 132.4/day | 2025-05-03 | CC0-1.0 | ai-agents, claude-code, codex | 2026-07-16 |
+| [mksglu/context-mode](https://github.com/mksglu/context-mode) | 18,992 | — | 132.1/day | 2026-02-23 | NOASSERTION | codex, recent-mcp | 2026-07-16 |
+| [langgenius/dify](https://github.com/langgenius/dify) | 149,028 | — | 125.1/day | 2023-04-12 | NOASSERTION | mcp | 2026-07-16 |
+| [nexu-io/html-anything](https://github.com/nexu-io/html-anything) | 7,796 | — | 117.1/day | 2026-05-11 | Apache-2.0 | agent-skills, recent-agent-skills | 2026-07-16 |
+| [diegosouzapw/OmniRoute](https://github.com/diegosouzapw/OmniRoute) | 17,910 | — | 116.7/day | 2026-02-13 | MIT | codex, recent-mcp | 2026-07-16 |
+| [microsoft/ai-agents-for-beginners](https://github.com/microsoft/ai-agents-for-beginners) | 69,475 | — | 116.7/day | 2024-11-28 | MIT | ai-agents | 2026-07-16 |
+| [decolua/9router](https://github.com/decolua/9router) | 22,356 | — | 115.9/day | 2026-01-05 | MIT | codex | 2026-07-16 |
+| [NVIDIA/SkillSpector](https://github.com/NVIDIA/SkillSpector) | 13,274 | — | 112.5/day | 2026-03-21 | Apache-2.0 | recent-agent-skills | 2026-07-16 |
+| [router-for-me/CLIProxyAPI](https://github.com/router-for-me/CLIProxyAPI) | 42,334 | — | 111.4/day | 2025-07-01 | MIT | claude-code, codex | 2026-07-16 |
+| [hesreallyhim/awesome-claude-code](https://github.com/hesreallyhim/awesome-claude-code) | 50,137 | — | 110.6/day | 2025-04-19 | NOASSERTION | agent-skills, claude-code | 2026-07-16 |
+| [agentskills/agentskills](https://github.com/agentskills/agentskills) | 23,143 | — | 109.0/day | 2025-12-16 | Apache-2.0 | agent-skills | 2026-07-16 |
+| [D4Vinci/Scrapling](https://github.com/D4Vinci/Scrapling) | 69,734 | — | 108.8/day | 2024-10-13 | BSD-3-Clause | mcp | 2026-07-16 |
+| [VoltAgent/awesome-agent-skills](https://github.com/VoltAgent/awesome-agent-skills) | 28,201 | — | 107.8/day | 2025-10-28 | MIT | agent-skills, ai-agents | 2026-07-16 |
+| [wanshuiyin/Auto-claude-code-research-in-sleep](https://github.com/wanshuiyin/Auto-claude-code-research-in-sleep) | 13,475 | — | 104.7/day | 2026-03-10 | MIT | codex, recent-agent-skills | 2026-07-16 |
+| [langchain-ai/langchain](https://github.com/langchain-ai/langchain) | 141,904 | — | 103.7/day | 2022-10-17 | MIT | ai-agents | 2026-07-16 |
+| [earthtojake/text-to-cad](https://github.com/earthtojake/text-to-cad) | 8,183 | — | 95.2/day | 2026-04-22 | MIT | recent-agent-skills | 2026-07-16 |
+| [AgriciDaniel/claude-obsidian](https://github.com/AgriciDaniel/claude-obsidian) | 9,472 | — | 94.1/day | 2026-04-07 | MIT | agent-skills | 2026-07-16 |
+| [infiniflow/ragflow](https://github.com/infiniflow/ragflow) | 85,193 | — | 89.9/day | 2023-12-12 | Apache-2.0 | ai-agents | 2026-07-16 |
+| [iOfficeAI/AionUi](https://github.com/iOfficeAI/AionUi) | 30,197 | — | 87.9/day | 2025-08-07 | Apache-2.0 | claude-code, codex | 2026-07-16 |
+| [greensock/gsap-skills](https://github.com/greensock/gsap-skills) | 11,603 | — | 86.4/day | 2026-03-04 | MIT | recent-agent-skills | 2026-07-16 |
+| [alirezarezvani/claude-skills](https://github.com/alirezarezvani/claude-skills) | 22,691 | — | 83.8/day | 2025-10-19 | MIT | agent-skills | 2026-07-16 |
+| [daytonaio/daytona](https://github.com/daytonaio/daytona) | 72,264 | — | 81.0/day | 2024-02-06 | NOASSERTION | ai-agents | 2026-07-16 |
+| [davila7/claude-code-templates](https://github.com/davila7/claude-code-templates) | 29,592 | — | 78.3/day | 2025-07-04 | MIT | claude-code | 2026-07-16 |
+| [n8n-io/n8n](https://github.com/n8n-io/n8n) | 196,635 | — | 76.2/day | 2019-06-22 | NOASSERTION | mcp | 2026-07-16 |
+| [steipete/CodexBar](https://github.com/steipete/CodexBar) | 18,426 | — | 76.0/day | 2025-11-16 | MIT | codex | 2026-07-16 |
+| [jlcodes99/cockpit-tools](https://github.com/jlcodes99/cockpit-tools) | 13,590 | — | 74.8/day | 2026-01-16 | NOASSERTION | codex | 2026-07-16 |
+| [revfactory/harness](https://github.com/revfactory/harness) | 8,365 | — | 74.6/day | 2026-03-26 | Apache-2.0 | recent-agent-skills | 2026-07-16 |
+| [aaif-goose/goose](https://github.com/aaif-goose/goose) | 51,276 | — | 74.1/day | 2024-08-23 | Apache-2.0 | ai-agents, mcp | 2026-07-16 |
+| [VectifyAI/PageIndex](https://github.com/VectifyAI/PageIndex) | 34,054 | — | 72.2/day | 2025-04-01 | MIT | ai-agents | 2026-07-16 |
+| [AgriciDaniel/claude-seo](https://github.com/AgriciDaniel/claude-seo) | 11,521 | — | 72.2/day | 2026-02-07 | MIT | recent-agent-skills | 2026-07-16 |
+| [bytedance/UI-TARS-desktop](https://github.com/bytedance/UI-TARS-desktop) | 37,999 | — | 69.9/day | 2025-01-19 | Apache-2.0 | mcp | 2026-07-16 |
+| [lobehub/lobehub](https://github.com/lobehub/lobehub) | 79,907 | — | 69.3/day | 2023-05-21 | NOASSERTION | mcp | 2026-07-16 |
+| [alibaba/zvec](https://github.com/alibaba/zvec) | 14,952 | — | 66.8/day | 2025-12-05 | Apache-2.0 | agent-skills | 2026-07-16 |
+| [YishenTu/claudian](https://github.com/YishenTu/claudian) | 14,092 | — | 63.1/day | 2025-12-05 | MIT | codex | 2026-07-16 |
+| [WenyuChiou/awesome-agentic-ai-zh](https://github.com/WenyuChiou/awesome-agentic-ai-zh) | 4,563 | — | 62.4/day | 2026-05-04 | MIT | model-context-protocol | 2026-07-16 |
+| [slopus/happy](https://github.com/slopus/happy) | 22,663 | — | 62.3/day | 2025-07-18 | MIT | codex | 2026-07-16 |
+| [CherryHQ/cherry-studio](https://github.com/CherryHQ/cherry-studio) | 48,647 | — | 62.1/day | 2024-05-24 | AGPL-3.0 | agent-skills, claude-code, codex | 2026-07-16 |
+| [nicobailon/visual-explainer](https://github.com/nicobailon/visual-explainer) | 9,254 | — | 61.6/day | 2026-02-16 | MIT | recent-agent-skills | 2026-07-16 |
+| [ashishpatel26/500-AI-Agents-Projects](https://github.com/ashishpatel26/500-AI-Agents-Projects) | 34,539 | — | 60.2/day | 2024-12-20 | MIT | ai-agents | 2026-07-16 |
+| [HKUDS/OpenSpace](https://github.com/HKUDS/OpenSpace) | 6,749 | — | 58.9/day | 2026-03-24 | MIT | recent-agent-skills | 2026-07-16 |
+| [jnMetaCode/superpowers-zh](https://github.com/jnMetaCode/superpowers-zh) | 6,947 | — | 58.4/day | 2026-03-20 | MIT | agent-skills, recent-agent-skills | 2026-07-16 |
+| [tt-a1i/archify](https://github.com/tt-a1i/archify) | 5,385 | — | 58.0/day | 2026-04-15 | MIT | recent-agent-skills | 2026-07-16 |
+| [patchy631/ai-engineering-hub](https://github.com/patchy631/ai-engineering-hub) | 36,525 | — | 57.7/day | 2024-10-21 | MIT | mcp | 2026-07-16 |
+| [aden-hive/hive](https://github.com/aden-hive/hive) | 10,707 | — | 57.6/day | 2026-01-12 | Apache-2.0 | agent-skills | 2026-07-16 |
+| [anysearch-ai/anysearch-skill](https://github.com/anysearch-ai/anysearch-skill) | 4,454 | — | 57.2/day | 2026-04-30 | Apache-2.0 | recent-agent-skills | 2026-07-16 |
+| [crewAIInc/crewAI](https://github.com/crewAIInc/crewAI) | 55,613 | — | 56.0/day | 2023-10-27 | MIT | ai-agents | 2026-07-16 |
+| [Mintplex-Labs/anything-llm](https://github.com/Mintplex-Labs/anything-llm) | 63,377 | — | 55.6/day | 2023-06-04 | MIT | ai-agents | 2026-07-16 |
+| [MinishLab/semble](https://github.com/MinishLab/semble) | 5,629 | — | 55.5/day | 2026-04-06 | MIT | model-context-protocol | 2026-07-16 |
+| [mem0ai/mem0](https://github.com/mem0ai/mem0) | 60,960 | — | 54.3/day | 2023-06-20 | Apache-2.0 | ai-agents | 2026-07-16 |
+| [Snailclimb/JavaGuide](https://github.com/Snailclimb/JavaGuide) | 157,070 | — | 52.5/day | 2018-05-07 | Apache-2.0 | mcp | 2026-07-16 |
+| [PDFMathTranslate/PDFMathTranslate](https://github.com/PDFMathTranslate/PDFMathTranslate) | 35,605 | — | 52.5/day | 2024-09-06 | AGPL-3.0 | mcp | 2026-07-16 |
+| [google/agents-cli](https://github.com/google/agents-cli) | 5,162 | — | 52.0/day | 2026-04-08 | Apache-2.0 | recent-agent-skills | 2026-07-16 |
+| [outsourc-e/hermes-workspace](https://github.com/outsourc-e/hermes-workspace) | 6,091 | — | 49.6/day | 2026-03-16 | MIT | recent-agent-skills | 2026-07-16 |
+| [butterbase-ai/butterbase](https://github.com/butterbase-ai/butterbase) | 2,763 | — | 47.6/day | 2026-05-20 | Apache-2.0 | recent-mcp | 2026-07-16 |
+| [holaboss-ai/holaOS](https://github.com/holaboss-ai/holaOS) | 5,498 | — | 47.1/day | 2026-03-22 | NOASSERTION | model-context-protocol | 2026-07-16 |
+| [AgriciDaniel/claude-ads](https://github.com/AgriciDaniel/claude-ads) | 7,099 | — | 45.7/day | 2026-02-11 | MIT | agent-skills | 2026-07-16 |
+| [xixu-me/awesome-persona-distill-skills](https://github.com/xixu-me/awesome-persona-distill-skills) | 4,566 | — | 45.1/day | 2026-04-06 | CC0-1.0 | recent-agent-skills | 2026-07-16 |
+| [PrefectHQ/fastmcp](https://github.com/PrefectHQ/fastmcp) | 26,231 | — | 44.2/day | 2024-11-30 | Apache-2.0 | model-context-protocol | 2026-07-16 |
+| [tigicion/dao-code](https://github.com/tigicion/dao-code) | 1,616 | — | 41.8/day | 2026-06-08 | MIT | recent-mcp | 2026-07-16 |
+| [Fincept-Corporation/FinceptTerminal](https://github.com/Fincept-Corporation/FinceptTerminal) | 28,500 | — | 41.5/day | 2024-08-29 | NOASSERTION | ai-agents | 2026-07-16 |
+| [eigent-ai/eigent](https://github.com/eigent-ai/eigent) | 14,570 | — | 41.4/day | 2025-07-29 | Apache-2.0 | agent-skills | 2026-07-16 |
+| [0xNyk/awesome-hermes-agent](https://github.com/0xNyk/awesome-hermes-agent) | 4,723 | — | 40.8/day | 2026-03-23 | NOASSERTION | recent-agent-skills | 2026-07-16 |
+| [tradesdontlie/tradingview-mcp](https://github.com/tradesdontlie/tradingview-mcp) | 4,412 | — | 40.3/day | 2026-03-29 | NOASSERTION | recent-mcp | 2026-07-16 |
+| [mudler/LocalAI](https://github.com/mudler/LocalAI) | 47,561 | — | 39.1/day | 2023-03-18 | MIT | mcp | 2026-07-16 |
+| [AsyncFuncAI/deepwiki-open](https://github.com/AsyncFuncAI/deepwiki-open) | 17,311 | — | 39.1/day | 2025-04-30 | MIT | codex | 2026-07-16 |
+| [op7418/CodePilot](https://github.com/op7418/CodePilot) | 6,136 | — | 38.2/day | 2026-02-06 | NOASSERTION | recent-agent-skills, recent-mcp | 2026-07-16 |
+| [Gentleman-Programming/engram](https://github.com/Gentleman-Programming/engram) | 5,479 | — | 36.4/day | 2026-02-16 | MIT | recent-mcp | 2026-07-16 |
+| [NousResearch/hermes-agent-self-evolution](https://github.com/NousResearch/hermes-agent-self-evolution) | 4,692 | — | 36.2/day | 2026-03-09 | NOASSERTION | recent-agent-skills | 2026-07-16 |
+| [microsoft/mcp-for-beginners](https://github.com/microsoft/mcp-for-beginners) | 16,762 | — | 35.8/day | 2025-04-04 | MIT | model-context-protocol | 2026-07-16 |
+| [epiral/bb-browser](https://github.com/epiral/bb-browser) | 5,973 | — | 35.8/day | 2026-01-31 | MIT | recent-mcp | 2026-07-16 |
+| [deanpeters/Product-Manager-Skills](https://github.com/deanpeters/Product-Manager-Skills) | 5,789 | — | 35.8/day | 2026-02-05 | NOASSERTION | recent-agent-skills | 2026-07-16 |
+| [langchain-ai/langgraph](https://github.com/langchain-ai/langgraph) | 37,417 | — | 34.9/day | 2023-08-09 | MIT | ai-agents | 2026-07-16 |
+| [Mouseww/anything-analyzer](https://github.com/Mouseww/anything-analyzer) | 3,306 | — | 34.6/day | 2026-04-12 | NOASSERTION | recent-mcp | 2026-07-16 |
+| [ComposioHQ/composio](https://github.com/ComposioHQ/composio) | 29,252 | — | 33.5/day | 2024-02-23 | MIT | ai-agents, mcp | 2026-07-16 |
+| [chatboxai/chatbox](https://github.com/chatboxai/chatbox) | 41,021 | — | 33.4/day | 2023-03-06 | GPL-3.0 | claude-code | 2026-07-16 |
+| [liyupi/ai-guide](https://github.com/liyupi/ai-guide) | 17,311 | — | 33.4/day | 2025-02-13 | NOASSERTION | codex | 2026-07-16 |
+| [danny-avila/LibreChat](https://github.com/danny-avila/LibreChat) | 40,803 | — | 32.6/day | 2023-02-12 | MIT | mcp | 2026-07-16 |
+| [twostraws/SwiftUI-Agent-Skill](https://github.com/twostraws/SwiftUI-Agent-Skill) | 4,304 | — | 32.3/day | 2026-03-05 | MIT | recent-agent-skills | 2026-07-16 |
+| [zhayujie/CowAgent](https://github.com/zhayujie/CowAgent) | 46,004 | — | 32.0/day | 2022-08-07 | MIT | ai-agents, claude-code, codex, mcp | 2026-07-16 |
+| [repowise-dev/repowise](https://github.com/repowise-dev/repowise) | 3,655 | — | 31.6/day | 2026-03-23 | AGPL-3.0 | recent-mcp | 2026-07-16 |
+| [excalidraw/excalidraw-mcp](https://github.com/excalidraw/excalidraw-mcp) | 4,934 | — | 30.4/day | 2026-02-04 | NOASSERTION | recent-mcp | 2026-07-16 |
+| [coleam00/excalidraw-diagram-skill](https://github.com/coleam00/excalidraw-diagram-skill) | 4,114 | — | 29.9/day | 2026-03-01 | NOASSERTION | recent-agent-skills | 2026-07-16 |
+| [jgraph/drawio-mcp](https://github.com/jgraph/drawio-mcp) | 4,829 | — | 29.4/day | 2026-02-02 | Apache-2.0 | recent-mcp | 2026-07-16 |
+| [ai-boost/awesome-harness-engineering](https://github.com/ai-boost/awesome-harness-engineering) | 3,094 | — | 28.3/day | 2026-03-29 | NOASSERTION | recent-mcp | 2026-07-16 |
+| [yvgude/lean-ctx](https://github.com/yvgude/lean-ctx) | 3,257 | — | 28.3/day | 2026-03-23 | Apache-2.0 | recent-mcp | 2026-07-16 |
+| [browser-act/skills](https://github.com/browser-act/skills) | 4,446 | — | 27.7/day | 2026-02-06 | MIT | recent-agent-skills | 2026-07-16 |
+| [AstrBotDevs/AstrBot](https://github.com/AstrBotDevs/AstrBot) | 36,423 | — | 27.7/day | 2022-12-08 | AGPL-3.0 | mcp | 2026-07-16 |
+| [tech-leads-club/agent-skills](https://github.com/tech-leads-club/agent-skills) | 4,877 | — | 27.3/day | 2026-01-19 | NOASSERTION | recent-agent-skills | 2026-07-16 |
+| [agno-agi/agno](https://github.com/agno-agi/agno) | 41,190 | — | 26.8/day | 2022-05-04 | Apache-2.0 | ai-agents | 2026-07-16 |
+| [topoteretes/cognee](https://github.com/topoteretes/cognee) | 27,970 | — | 26.3/day | 2023-08-16 | Apache-2.0 | agent-skills | 2026-07-16 |
+| [CoplayDev/unity-mcp](https://github.com/CoplayDev/unity-mcp) | 12,531 | — | 25.8/day | 2025-03-18 | MIT | model-context-protocol | 2026-07-16 |
+| [Unclecheng-li/VulnClaw](https://github.com/Unclecheng-li/VulnClaw) | 2,096 | — | 23.4/day | 2026-04-18 | MIT | recent-mcp | 2026-07-16 |
+| [thedaviddias/Front-End-Checklist](https://github.com/thedaviddias/Front-End-Checklist) | 73,222 | — | 22.9/day | 2017-10-16 | NOASSERTION | ai-agents | 2026-07-16 |
+| [aws/agent-toolkit-for-aws](https://github.com/aws/agent-toolkit-for-aws) | 1,884 | — | 22.3/day | 2026-04-23 | Apache-2.0 | recent-mcp | 2026-07-16 |
+| [giancarloerra/SocratiCode](https://github.com/giancarloerra/SocratiCode) | 3,118 | — | 22.2/day | 2026-02-26 | AGPL-3.0 | recent-mcp | 2026-07-16 |
+| [TabularisDB/tabularis](https://github.com/TabularisDB/tabularis) | 3,724 | — | 21.8/day | 2026-01-26 | Apache-2.0 | recent-mcp | 2026-07-16 |
+| [mcp-use/mcp-use](https://github.com/mcp-use/mcp-use) | 10,317 | — | 21.7/day | 2025-03-28 | MIT | model-context-protocol | 2026-07-16 |
+| [ZenNotes/zennotes](https://github.com/ZenNotes/zennotes) | 2,031 | — | 21.1/day | 2026-04-11 | MIT | recent-mcp | 2026-07-16 |
+| [siyuan-note/siyuan](https://github.com/siyuan-note/siyuan) | 45,145 | — | 21.0/day | 2020-08-30 | AGPL-3.0 | mcp | 2026-07-16 |
+| [chroma-core/chroma](https://github.com/chroma-core/chroma) | 28,805 | — | 20.9/day | 2022-10-05 | Apache-2.0 | ai-agents | 2026-07-16 |
+| [breaking-brake/cc-wf-studio](https://github.com/breaking-brake/cc-wf-studio) | 5,321 | — | 20.7/day | 2025-11-01 | NOASSERTION | agent-skills | 2026-07-16 |
+| [Observal/Observal](https://github.com/Observal/Observal) | 2,191 | — | 20.2/day | 2026-03-30 | Apache-2.0 | recent-mcp | 2026-07-16 |
+| [anysearch-ai/anysearch-mcp-server](https://github.com/anysearch-ai/anysearch-mcp-server) | 1,541 | — | 19.8/day | 2026-04-30 | Apache-2.0 | recent-mcp | 2026-07-16 |
+| [louislva/claude-peers-mcp](https://github.com/louislva/claude-peers-mcp) | 2,164 | — | 18.4/day | 2026-03-21 | MIT | recent-mcp | 2026-07-16 |
+| [knowsuchagency/mcp2cli](https://github.com/knowsuchagency/mcp2cli) | 2,300 | — | 17.7/day | 2026-03-09 | MIT | recent-mcp | 2026-07-16 |
+| [UsefulSoftwareCo/executor](https://github.com/UsefulSoftwareCo/executor) | 2,703 | — | 17.0/day | 2026-02-07 | MIT | recent-mcp | 2026-07-16 |
+| [jeecgboot/JeecgBoot](https://github.com/jeecgboot/JeecgBoot) | 47,098 | — | 16.9/day | 2018-11-26 | Apache-2.0 | claude-code, codex, mcp | 2026-07-16 |
+| [netdata/netdata](https://github.com/netdata/netdata) | 79,657 | — | 16.7/day | 2013-06-17 | GPL-3.0 | mcp | 2026-07-16 |
+| [didilili/ai-agents-from-zero](https://github.com/didilili/ai-agents-from-zero) | 2,806 | — | 16.6/day | 2026-01-29 | MIT | recent-mcp | 2026-07-16 |
+| [moltis-org/moltis](https://github.com/moltis-org/moltis) | 2,780 | — | 16.5/day | 2026-01-29 | MIT | recent-mcp | 2026-07-16 |
+| [iFurySt/open-codex-computer-use](https://github.com/iFurySt/open-codex-computer-use) | 1,452 | — | 16.0/day | 2026-04-17 | MIT | model-context-protocol | 2026-07-16 |
+| [Mcp-Brasil/mcp-brasil](https://github.com/Mcp-Brasil/mcp-brasil) | 1,678 | — | 14.9/day | 2026-03-26 | MIT | recent-mcp | 2026-07-16 |
+| [rohitg00/awesome-claude-code-toolkit](https://github.com/rohitg00/awesome-claude-code-toolkit) | 2,332 | — | 14.4/day | 2026-02-04 | Apache-2.0 | recent-mcp | 2026-07-16 |
+| [forloopcodes/contextplus](https://github.com/forloopcodes/contextplus) | 1,955 | — | 14.0/day | 2026-02-27 | MIT | recent-mcp | 2026-07-16 |
+| [mindsdb/mindshub](https://github.com/mindsdb/mindshub) | 39,441 | — | 13.6/day | 2018-08-02 | MIT | codex, mcp | 2026-07-16 |
+| [0xMassi/webclaw](https://github.com/0xMassi/webclaw) | 1,731 | — | 13.5/day | 2026-03-10 | AGPL-3.0 | recent-mcp | 2026-07-16 |
+| [blitzdotdev/blitz-mac](https://github.com/blitzdotdev/blitz-mac) | 1,723 | — | 13.4/day | 2026-03-10 | Apache-2.0 | recent-mcp | 2026-07-16 |
+| [6551Team/opennews-mcp](https://github.com/6551Team/opennews-mcp) | 1,872 | — | 13.3/day | 2026-02-26 | MIT | recent-mcp | 2026-07-16 |
+| [xixu-me/xget](https://github.com/xixu-me/xget) | 8,152 | — | 13.1/day | 2024-11-01 | AGPL-3.0 | agent-skills | 2026-07-16 |
+| [homeassistant-ai/ha-mcp](https://github.com/homeassistant-ai/ha-mcp) | 3,998 | — | 13.1/day | 2025-09-14 | MIT | model-context-protocol | 2026-07-16 |
+| [jgravelle/jcodemunch-mcp](https://github.com/jgravelle/jcodemunch-mcp) | 2,020 | — | 12.8/day | 2026-02-09 | NOASSERTION | model-context-protocol, recent-mcp | 2026-07-16 |
+| [kunchenguid/axi](https://github.com/kunchenguid/axi) | 1,496 | — | 12.8/day | 2026-03-21 | MIT | recent-mcp | 2026-07-16 |
+| [getsentry/XcodeBuildMCP](https://github.com/getsentry/XcodeBuildMCP) | 6,076 | — | 12.3/day | 2025-03-09 | MIT | model-context-protocol | 2026-07-16 |
+| [ThinkInAIXYZ/deepchat](https://github.com/ThinkInAIXYZ/deepchat) | 6,106 | — | 11.8/day | 2025-02-14 | Apache-2.0 | agent-skills | 2026-07-16 |
+| [can4hou6joeng4/boss-agent-cli](https://github.com/can4hou6joeng4/boss-agent-cli) | 1,389 | — | 11.7/day | 2026-03-20 | MIT | model-context-protocol | 2026-07-16 |
+| [phodal/routa](https://github.com/phodal/routa) | 1,743 | — | 11.5/day | 2026-02-16 | MIT | recent-mcp | 2026-07-16 |
+| [alibaba/nacos](https://github.com/alibaba/nacos) | 33,166 | — | 11.2/day | 2018-06-15 | Apache-2.0 | mcp | 2026-07-16 |
+| [chrisryugj/korean-law-mcp](https://github.com/chrisryugj/korean-law-mcp) | 2,222 | — | 10.6/day | 2025-12-19 | MIT | model-context-protocol | 2026-07-16 |
+| [atilaahmettaner/tradingview-mcp](https://github.com/atilaahmettaner/tradingview-mcp) | 3,550 | — | 10.4/day | 2025-08-08 | MIT | model-context-protocol | 2026-07-16 |
+| [Kong/kong](https://github.com/Kong/kong) | 43,793 | — | 10.3/day | 2014-11-17 | Apache-2.0 | mcp | 2026-07-16 |
+| [Upsonic/Upsonic](https://github.com/Upsonic/Upsonic) | 7,909 | — | 10.1/day | 2024-05-26 | MIT | model-context-protocol | 2026-07-16 |
+| [jau123/MeiGen-AI-Design-MCP](https://github.com/jau123/MeiGen-AI-Design-MCP) | 1,577 | — | 9.9/day | 2026-02-07 | MIT | model-context-protocol, recent-mcp | 2026-07-16 |
+| [xu-xiang/everything-claude-code-zh](https://github.com/xu-xiang/everything-claude-code-zh) | 1,668 | — | 9.8/day | 2026-01-27 | MIT | recent-mcp | 2026-07-16 |
+| [IBM/mcp-context-forge](https://github.com/IBM/mcp-context-forge) | 4,101 | — | 9.4/day | 2025-05-08 | Apache-2.0 | model-context-protocol | 2026-07-16 |
+| [bethington/ghidra-mcp](https://github.com/bethington/ghidra-mcp) | 2,826 | — | 8.8/day | 2025-08-30 | Apache-2.0 | model-context-protocol | 2026-07-16 |
+| [IvanMurzak/Unity-MCP](https://github.com/IvanMurzak/Unity-MCP) | 3,568 | — | 7.6/day | 2025-04-02 | Apache-2.0 | model-context-protocol | 2026-07-16 |
+| [tractorjuice/arc-kit](https://github.com/tractorjuice/arc-kit) | 2,088 | — | 7.6/day | 2025-10-14 | NOASSERTION | model-context-protocol | 2026-07-16 |
+| [universal-tool-calling-protocol/code-mode](https://github.com/universal-tool-calling-protocol/code-mode) | 1,514 | — | 6.1/day | 2025-11-11 | MPL-2.0 | model-context-protocol | 2026-07-16 |
+| [stickerdaniel/linkedin-mcp-server](https://github.com/stickerdaniel/linkedin-mcp-server) | 2,795 | — | 6.1/day | 2025-04-13 | Apache-2.0 | model-context-protocol | 2026-07-16 |
+| [jamubc/gemini-mcp-tool](https://github.com/jamubc/gemini-mcp-tool) | 2,255 | — | 5.9/day | 2025-06-29 | NOASSERTION | model-context-protocol | 2026-07-16 |
+| [nanbingxyz/5ire](https://github.com/nanbingxyz/5ire) | 5,275 | — | 5.7/day | 2024-01-06 | NOASSERTION | model-context-protocol | 2026-07-16 |
+| [kubesphere/kubesphere](https://github.com/kubesphere/kubesphere) | 17,002 | — | 5.7/day | 2018-04-21 | NOASSERTION | agent-skills | 2026-07-16 |
+| [zinja-coder/jadx-ai-mcp](https://github.com/zinja-coder/jadx-ai-mcp) | 2,493 | — | 5.4/day | 2025-04-07 | Apache-2.0 | model-context-protocol | 2026-07-16 |
+| [ikaijua/Awesome-AITools](https://github.com/ikaijua/Awesome-AITools) | 6,087 | — | 5.1/day | 2023-04-02 | NOASSERTION | agent-skills | 2026-07-16 |
+| [zcaceres/markdownify-mcp](https://github.com/zcaceres/markdownify-mcp) | 2,840 | — | 4.9/day | 2024-12-18 | MIT | model-context-protocol | 2026-07-16 |
+| [metatool-ai/metamcp](https://github.com/metatool-ai/metamcp) | 2,527 | — | 4.7/day | 2025-01-22 | MIT | model-context-protocol | 2026-07-16 |
+| [IBM/AssetOpsBench](https://github.com/IBM/AssetOpsBench) | 2,025 | — | 4.6/day | 2025-05-01 | Apache-2.0 | model-context-protocol | 2026-07-16 |
+| [vibheksoni/stealth-browser-mcp](https://github.com/vibheksoni/stealth-browser-mcp) | 1,516 | — | 4.4/day | 2025-08-09 | MIT | model-context-protocol | 2026-07-16 |
+| [gosom/google-maps-scraper](https://github.com/gosom/google-maps-scraper) | 4,923 | — | 4.2/day | 2023-04-22 | MIT | agent-skills | 2026-07-16 |
+| [stacklok/toolhive](https://github.com/stacklok/toolhive) | 1,945 | — | 4.0/day | 2025-03-12 | Apache-2.0 | model-context-protocol | 2026-07-16 |
+| [CoderGamester/mcp-unity](https://github.com/CoderGamester/mcp-unity) | 1,829 | — | 3.7/day | 2025-03-13 | MIT | model-context-protocol | 2026-07-16 |
+| [trpc-group/trpc-agent-go](https://github.com/trpc-group/trpc-agent-go) | 1,551 | — | 3.6/day | 2025-05-14 | Apache-2.0 | model-context-protocol | 2026-07-16 |
+| [metalbear-co/mirrord](https://github.com/metalbear-co/mirrord) | 5,219 | — | 3.2/day | 2022-02-01 | MIT | agent-skills | 2026-07-16 |
+| [tinyplex/tinybase](https://github.com/tinyplex/tinybase) | 5,123 | — | 3.1/day | 2021-12-31 | MIT | agent-skills | 2026-07-16 |
+| [the-open-agent/openagent](https://github.com/the-open-agent/openagent) | 5,407 | — | 2.4/day | 2020-05-29 | Apache-2.0 | model-context-protocol | 2026-07-16 |
+| [designcomputer/mysql_mcp_server](https://github.com/designcomputer/mysql_mcp_server) | 1,331 | — | 2.3/day | 2024-12-03 | MIT | model-context-protocol | 2026-07-16 |
+
+## Reviewed catalog repositories seen today
+
+| Repository | Stars | Matched searches |
+| --- | ---: | --- |
+| [DietrichGebert/ponytail](https://github.com/DietrichGebert/ponytail) | 84,342 | agent-skills, ai-agents, claude-code |
+| [anthropics/skills](https://github.com/anthropics/skills) | 161,573 | agent-skills |
+| [addyosmani/agent-skills](https://github.com/addyosmani/agent-skills) | 78,675 | agent-skills, claude-code, codex, recent-agent-skills |
+| [nextlevelbuilder/ui-ux-pro-max-skill](https://github.com/nextlevelbuilder/ui-ux-pro-max-skill) | 106,345 | claude-code, codex |
+| [phuryn/pm-skills](https://github.com/phuryn/pm-skills) | 23,801 | agent-skills, recent-agent-skills |
+| [ChromeDevTools/chrome-devtools-mcp](https://github.com/ChromeDevTools/chrome-devtools-mcp) | 47,062 | mcp |
+| [OthmanAdi/planning-with-files](https://github.com/OthmanAdi/planning-with-files) | 25,405 | agent-skills, codex |
+| [JimLiu/baoyu-skills](https://github.com/JimLiu/baoyu-skills) | 23,704 | agent-skills |
+| [upstash/context7](https://github.com/upstash/context7) | 59,185 | mcp |
+| [K-Dense-AI/scientific-agent-skills](https://github.com/K-Dense-AI/scientific-agent-skills) | 30,999 | agent-skills |
+| [wshobson/agents](https://github.com/wshobson/agents) | 37,951 | agent-skills, ai-agents, claude-code, mcp |
+| [github/awesome-copilot](https://github.com/github/awesome-copilot) | 36,650 | agent-skills |
+| [microsoft/playwright-mcp](https://github.com/microsoft/playwright-mcp) | 35,149 | mcp |
+| [github/github-mcp-server](https://github.com/github/github-mcp-server) | 31,481 | mcp |
+| [oraios/serena](https://github.com/oraios/serena) | 26,510 | codex |
+| [Agents365-ai/drawio-skill](https://github.com/Agents365-ai/drawio-skill) | 5,852 | agent-skills |
+| [trailofbits/skills](https://github.com/trailofbits/skills) | 6,141 | agent-skills |
+| [antfu/skills](https://github.com/antfu/skills) | 5,604 | agent-skills, recent-agent-skills |
+| [dotnet/skills](https://github.com/dotnet/skills) | 4,633 | recent-agent-skills |
+| [firecrawl/firecrawl-mcp-server](https://github.com/firecrawl/firecrawl-mcp-server) | 6,967 | model-context-protocol |
+| [exa-labs/exa-mcp-server](https://github.com/exa-labs/exa-mcp-server) | 4,728 | model-context-protocol |
+| [taylorwilsdon/google_workspace_mcp](https://github.com/taylorwilsdon/google_workspace_mcp) | 2,856 | model-context-protocol |
+| [blazickjp/arxiv-mcp-server](https://github.com/blazickjp/arxiv-mcp-server) | 2,961 | model-context-protocol |
+
+## Search evidence
+
+- **MCP** (40 API results): `topic:mcp archived:false fork:false pushed:>=2026-05-17`
+- **Model Context Protocol** (40 API results): `topic:model-context-protocol archived:false fork:false pushed:>=2026-05-17`
+- **Agent skills** (40 API results): `topic:agent-skills archived:false fork:false pushed:>=2026-05-17`
+- **Claude Code** (40 API results): `topic:claude-code archived:false fork:false pushed:>=2026-05-17`
+- **Codex** (40 API results): `topic:codex archived:false fork:false pushed:>=2026-05-17`
+- **AI agents** (40 API results): `topic:ai-agents archived:false fork:false pushed:>=2026-05-17`
+- **Recently created MCP repositories** (40 API results): `mcp in:name,description archived:false fork:false created:>=2026-01-17`
+- **Recently created agent-skill repositories** (40 API results): `skills agent in:name,description archived:false fork:false created:>=2026-01-17`
+
+The machine-readable evidence, query counts, bounded observation history, and catalog status are in [`catalog/discovered.json`](../catalog/discovered.json).

package/docs/EVALUATION_PROTOCOL.md

@@ -0,0 +1,40 @@
+# Evaluation protocol and uncertainty
+
+Loadout evaluates package **evidence**, not the universal quality of an agent.
+Scores are category-specific, reproducible, and carry an uncertainty statement.
+
+## Category 1: Skill package hygiene
+
+Inputs are a pinned source snapshot and discovered `SKILL.md` directories. Checks:
+
+- required frontmatter name and description parse;
+- names are unique within the package;
+- paths stay within the snapshot and are not symlinks;
+- instruction text is scanned for scripts, credential references, and external
+  domains; and
+- the package declares at least one compatible target layout.
+
+Result: `ready`, `needs-review`, `blocked`, or `not-applicable`, plus findings.
+An absent component category is `not-applicable` and never blocks a different
+supported component. This says nothing about whether the instructions improve a
+particular model's output.
+
+## Category 2: MCP manifest safety
+
+Inputs are static JSON manifests only. Checks:
+
+- valid command/argument structure or HTTPS endpoint;
+- no inline credential values;
+- declared environment variable names only;
+- no executable runs during evaluation; and
+- each server has a concrete compatible configuration target.
+
+Result: `ready`, `needs-approval`, or `blocked`, plus the exact permission/finding
+that led there. A passing result does not attest to server reliability or data access.
+
+## Uncertainty rules
+
+No score uses stars, social mentions, or a single benchmark as proof of quality.
+Missing evidence lowers confidence; it is not treated as negative evidence. Evaluators
+are deterministic over the same pinned snapshot and report their version and inputs.
+Human review remains mandatory for all findings that require approval.