Claude Code mod
agent-watcher
Logs every agent dispatch and forked skill with engine-measured usage rows; tracks pr-review / pr-team runs: Phase 5 band, matrix rows, bubble Stop guard
Install
/plugin marketplace add t968rs/operator-straps
/plugin install agent-watcher@operator-strapsMods install like plugins. Run /reload-plugins or start a new session to load it.
What it can access
- Reach
- L2Writes or runs: runs processes, writes files, reads files, reads env vars, reads mod state, draws
- Sees
- Agent calls, every tool call, classic hooks, Bash calls, PowerShell calls, every prompt, compaction
- Draws in
AbovePrompt- Hooks
tool.callturn.stepturn.completeclassic.Stopsession.startprompt.submitsession.endsession.compactui.render
From a static scan of the mod's code by awesome-claude-code-mods. It shows what the code asks for, not a security review.
How Claude Code mods work
- A mod is a plugin with a JavaScript or TypeScript module that runs inside Claude Code (version 2.1.287 or later). It can draw panes beside the transcript, a band above the prompt, buttons and toasts, restyle the spinner and tool rows, and watch or rewrite tool calls and prompts.
- It installs like any plugin: add its marketplace with /plugin marketplace add owner/repo, then /plugin install name@marketplace.
- Check what a mod can reach before installing: it runs with your permissions. Reach goes from drawing only, to reading files, to writing files or running processes, to the network. Run claude plugin validate on a clone to list its hooks and calls without running it.
- Turn one off from the Installed tab in /plugin, or start Claude Code with --safe-mode to skip every customization for a session.
Details
- Repository
- t968rs/operator-straps
- Folder
plugins/agent-watcher- Marketplace
operator-straps- Author
- Kevin Gabelman
- License
- MIT
- Last updated
- 7 November 2026
- Listed in
- awesome-claude-code-mods
FAQ
What is agent-watcher?
Logs every agent dispatch and forked skill with engine-measured usage rows; tracks pr-review / pr-team runs: Phase 5 band, matrix rows, bubble Stop guard It is a mod for Claude Code, from t968rs/operator-straps on GitHub.
What can the agent-watcher mod access?
Its widest reach is "writes or runs" (runs processes, writes files, reads files, reads env vars, reads mod state, draws). It sees Agent calls, every tool call, classic hooks, Bash calls, PowerShell calls, every prompt, compaction. It draws in AbovePrompt. Like any mod, it runs with your permissions.
How do I install agent-watcher?
Inside Claude Code (v2.1.287 or later), run /plugin marketplace add t968rs/operator-straps and then /plugin install agent-watcher@operator-straps.
More git & github mods
agent-history
zirkelc/agent-history
Agent History: exports every Claude Code session and its subagents to a folder, and commits and pushes it when the folder is a git repository
aidiff-gate
JohnRiceML/aidiff-gate
Keeps AI-written code honest: flags flaky and bloated tests, trims wall-of-text replies, and makes you write the 'why' before a commit. Uses TypeSafe's Jev as a fast judge.
agent-charter
boxiaolanya2008/claudecode-plus-mod
Pin an agent charter into the system prompt: skills-first workflow, per-language comment standards, commit format and branch discipline.
airlock
BillyMRX1/airlock
Routes supported project edits into a Git worktree for review, accept or reject. Worktree routing is not a filesystem sandbox.
afferent
Mark-Hickey/claude-nine
Senses upstream change: keeps ~/projects current, flags rule, tool and branch drift, and leaves the acting to but, chezmoi and the SOPs
autocommit
Iliesseu28/claude-autocommit
Turns every turn Claude takes into a clean git commit: only the files that turn changed, a Conventional Commits message written by Haiku, a secret scan before every commit, undo and squash. Plus a band above the prompt with model, effort, context and commits.
Browse the most popular Claude Code mods, all git & github mods, or the skills and MCP server directories.