claude-software-3-0-marketplace
CLAUDE.md2 starsChanged 5 months agoMITView on GitHub
Author/dev project memory. Not shipped context — adding this marketplace registers an index; nothing in this repo is loaded into an end user's Claude Code session. This repo is the odere-pro marketplace: a dedicated, registry-only aggregator. It ships one source of truth — .claude-plugin/marketplace.json — that lists each plugin by a github source. Each plugin lives in its own repo and installs individually as <plugin>@odere-pro. - .claude-plugin/marketplace.json — the manifest (name odere-pro; one github-source entry per plugin; no version, no…
In this repository11 other files it gives its agents.
- .claude/CLAUDE.mdCLAUDE.md
- .claude-plugin/CLAUDE.mdCLAUDE.md
- .claude/teams/CLAUDE.mdCLAUDE.md
- docs/CLAUDE.mdCLAUDE.md
- .github/CLAUDE.mdCLAUDE.md
- site/CLAUDE.mdCLAUDE.md
- add-pluginSkill
- list-pluginsSkill
- remove-pluginSkill
- update-pluginSkill
- vet-pluginSkill
The file
We split this CLAUDE.md into the rules it gives an agent
2 rules. About 0 are common, 0 are distinctive, and 2 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈0
Distinctive · estimate
This author's own choices. Worth a look.
2
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 35
Every navigational layer carries a CLAUDE.md that says *just enough* for its level.
- Very specific: line 40
Use the agent-driven skills (each vets, edits the manifest + README + CHANGELOG, runs gates, and opens a PR): /vet-plugin <repo> (read-only …
≈ “Common” is an estimate: at least 1 in 5 of the 27,641 CLAUDE.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
claude-software-3-0-marketplace — registry development
Author/dev project memory. Not shipped context — adding this marketplace registers an index; nothing in this repo is loaded into an end user's Claude Code session.
This repo is the odere-pro marketplace: a dedicated, registry-only aggregator. It ships one
source of truth — .claude-plugin/marketplace.json — that lists each plugin by a github source.
Each plugin lives in its own repo and installs individually as <plugin>@odere-pro.
What ships (the registry)
.claude-plugin/marketplace.json— the manifest (nameodere-pro; onegithub-source entry per plugin; noversion, nosha). The only payload. Deep detail:.claude-plugin/CLAUDE.md.README.md,LICENSE— what the marketplace is and how to use it.
This repo has no plugin.json and no plugin components (skills/, agents/, hooks/,
commands/). It is not a plugin.
What doesn't ship (author-only — governance, CI, harness)
| Path | Role | Detail |
.claude-plugin/ | the manifest (registry payload) | .claude-plugin/CLAUDE.md |
.claude/ | dev harness: settings, hooks, rules, the add-plugin skills + agent, the registry-dev team, statusline, MCP | .claude/CLAUDE.md |
.claude/teams/ | author-only spawnable agent teams (the registry-dev engineering team) | .claude/teams/CLAUDE.md |
.github/ | CI, supply-chain workflows, Dependabot, CODEOWNERS | .github/CLAUDE.md |
tests/gates/ | the validation gate suite | tests/gates/CLAUDE.md |
docs/ | human-readable references (adding plugins; the dev teams + roadmaps) | docs/CLAUDE.md |
| root docs | SOFTWARE-3-0.md, CONTRIBUTING.md, SECURITY.md, SUPPORT.md, CODE_OF_CONDUCT.md, CHANGELOG.md | governance |
| hygiene | .editorconfig, .gitattributes, .markdownlint*.jsonc, .prettierrc, package.json | repo config |
Nested CLAUDE.md convention
Every navigational layer carries a CLAUDE.md that says just enough for its level; detail deepens
as you descend. tests/gates/07-claude-md-coverage.sh enforces presence for the documented layers.
How to manage plugins
Use the agent-driven skills (each vets, edits the manifest + README + CHANGELOG, runs gates, and opens a
PR): /vet-plugin <repo> (read-only preflight), /add-plugin <repo>, /update-plugin <name>
(refresh / --repo repoint / --name rename), /remove-plugin <name>. Full process, contract, and
manual fallback: docs/adding-plugins.md,
CONTRIBUTING.md, and .claude-plugin/CLAUDE.md.
Gate map
bash tests/gates/run-all.sh before any PR (CI runs the same). G0 harness-integrity · G1 json-parses
· G2 marketplace-shape · G3 no-absolute-paths · G4 secret-scan · G5 doc-links · G6 shellcheck ·
G7 claude-md-coverage · G8 markdown-lint (advisory) · G9 readme-in-sync · G10 vet-verdict-schema.
See tests/gates/CLAUDE.md.
Verify before commit
bash tests/gates/run-all.sh # full gate suite is green (the authoritative check; CI runs this)
claude plugin validate . # manifest validatesWhile the registry is empty,
claude plugin validate --strictwarns "Marketplace has no plugins defined" (a warning--strictpromotes to failure). That's expected — the gate suite is the green check;--strictbecomes meaningful once the first plugin is listed via/add-plugin.
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

