Skip to content
Back to search

agents-hub

One source of truth for AI agent skills, subagents, MCP servers and global instructions (Claude Code, Codex, Cursor)

Safety scan

A static text scan, not a guarantee. Select the badge to see what was scanned and every finding.

Checked today
Badge for your README
findsafeskills safety rating

Markdown

[![findsafeskills safety rating](/api/listings/sujit-gholap-121--agents-hub/badge.svg)](/listings/sujit-gholap-121--agents-hub)

HTML

<a href="/listings/sujit-gholap-121--agents-hub"><img src="/api/listings/sujit-gholap-121--agents-hub/badge.svg" alt="findsafeskills safety rating" /></a>

Install

i

What was scanned

The files the safety scan read, as written by the repository's author and shown as plain text. Lines with a finding are marked.

README.md · README · 95 lines
  1. # agents-hub
  2. The single source of truth for my AI agent setup. Claude Code, Codex and Cursor all get their
  3. **skills, subagents, MCP servers and global instructions** from here, via one command:
  4. ```sh
  5. agents-sync
  6. ```
  7. Backed up to the public repo `sujit-gholap-121/agents-hub` (every sync commits and pushes).
  8. ## Layout
  9. ```
  10. ~/.agents/
  11. ├── skills/<name>/SKILL.md skills (flat, the open Agent Skills format)
  12. ├── agents/<name>.md subagents (Claude-style markdown + frontmatter)
  13. ├── mcp/servers.json MCP servers, with ${VAR} placeholders for secrets
  14. ├── instructions/AGENTS.md global rules for every agent
  15. ├── experimental/{skills,agents} only switched on with --with-experimental
  16. ├── groups.json which group each skill belongs to
  17. ├── groups/ generated browsing view (links), don't edit
  18. ├── bin/agents-sync the sync tool
  19. ├── .env secrets for servers.json (NOT in git, chmod 600)
  20. └── state/ what the tool manages, logs, config backups (NOT in git)
  21. ```
  22. ## What goes where
  23. | Hub | Claude Code | Codex | Cursor |
  24. |---|---|---|---|
  25. | skills | links in `~/.claude/skills` | reads `~/.agents/skills` directly | reads `~/.agents/skills` directly |
  26. | agents | links in `~/.claude/agents` | generated `~/.codex/agents/*.toml` | links in `~/.cursor/agents` (app) + per-project (CLI, see below) |
  27. | MCP | `claude mcp add-json -s user` | generated block in `~/.codex/config.toml` | `~/.cursor/mcp.json` |
  28. | instructions | `~/.claude/CLAUDE.md` (link) | `~/.codex/AGENTS.md` (link) | `~/.cursor/rules/agents-hub.mdc` (always-apply rule) |
  29. **Cursor CLI and agents:** `cursor-agent` only loads subagents from the project it is opened in.
  30. Register a project once and every sync keeps it current (its git ignores the links locally):
  31. ```sh
  32. agents-sync --project ~/Workspace/Github/escalara-portal
  33. agents-sync --unproject ~/Workspace/Github/escalara-portal
  34. ```
  35. ## How a sync works
  36. 1. **Import.** Anything new that an installer put straight into one app (`npx skills add`,
  37. the Agency Agents app, `claude mcp add`, a hand-made Codex agent) is moved into the hub.
  38. Secret-looking MCP values go to `.env` and become `${VAR}` placeholders.
  39. 2. **Normalize.** Agent descriptions over 160 characters are shortened so all of them fit Claude
  40. Code's 15k-token limit. The original text is kept in `full_description:`.
  41. 3. **Export.** The hub is written into each app. Only entries the tool manages are touched.
  42. App-owned items stay where they are: `~/.claude/skills/synced`, `~/.codex/skills/.system`,
  43. `~/.cursor/skills-cursor`, Codex plugin MCPs, and MCPs listed in `servers.json` → `unmanaged`.
  44. 4. **Group, commit, push.**
  45. ## Everyday use
  46. ```sh
  47. # new skill
  48. npx skills add <repo> # installs into ~/.agents/skills
  49. # or: mkdir ~/.agents/skills/my-skill && $EDITOR ~/.agents/skills/my-skill/SKILL.md
  50. # new agent
  51. $EDITOR ~/.agents/agents/engineering-my-agent.md # prefix = group
  52. # new MCP server: add to mcp/servers.json, e.g.
  53. # "github": { "command": "npx", "args": ["-y", "@modelcontextprotocol/server-github"],
  54. # "env": { "GITHUB_TOKEN": "${GITHUB_TOKEN}" }, "targets": ["claude", "codex", "cursor"] }
  55. # and put GITHUB_TOKEN=... in ~/.agents/.env
  56. # experiments
  57. mv my-skill ~/.agents/experimental/skills/ && agents-sync --with-experimental
  58. agents-sync # then restart the apps (they read these at startup)
  59. agents-sync --dry-run # show what would change, change nothing
  60. ```
  61. `targets` (default: all three) limits an MCP server to some apps. `"enabled": false` turns it off
  62. everywhere. `"extra": {"codex": {...}}` adds app-specific fields (for example Codex timeouts).
  63. ## New machine
  64. ```sh
  65. git clone https://github.com/sujit-gholap-121/agents-hub ~/.agents
  66. ln -s ~/.agents/bin/agents-sync ~/.local/bin/agents-sync
  67. # recreate ~/.agents/.env (secrets are never in git), then:
  68. agents-sync
  69. ```
  70. ## Undo
  71. Everything is in git: `git -C ~/.agents log`, then `git revert <commit>` and `agents-sync`.
  72. The pre-hub backups are `backup-skills-2026-11-07.tgz` and `backup-agents-2026-11-07.tgz`
  73. (not in git), and the last Codex config before each change is in `state/codex-config.toml.bak`.

Ratings and comments

Sign in to rate this skill.

Sign in

Free to use. If it’s helped, pass it on.