Updated 3d ago
by Cloud-Eye-Prime
mcp-agent-forge
Dispatch AI coding agents (Claude Code, Copilot, OpenCode) as MCP tools - with resumable sessions, a cost ledger, admission and work-jail guards, and zikr + auditor skills against drift
27 GitHub stars (27)
Steady at 27 stars — no new stars in the 3 days since Nov 19.
- 2 skills (2 skills found in its SKILL.md files)
- 2 forks (2 forks)
- 3d ago last update (Last updated 3d ago)
Install
Skills CLI
npx skills add Cloud-Eye-Prime/mcp-agent-forge
Asks which skills and agents to install. Just one: npx skills add Cloud-Eye-Prime/mcp-agent-forge --skill auditor — add -g for every project.
Claude Code
git clone https://github.com/Cloud-Eye-Prime/mcp-agent-forgecp -r "mcp-agent-forge/skills/auditor" ~/.claude/skills/
Copies the auditor skill into ~/.claude/skills/ for every project (use .claude/skills/ for one project). Repeat for the other skills listed below.
Plugin
/plugin marketplace add Cloud-Eye-Prime/mcp-agent-forge
Run inside Claude Code if the repository ships a plugin marketplace, then pick plugins from /plugin.
Skills can run code on your machine — read the SKILL.md first.
What it does
In the author's own words, from each skill's SKILL.md.
-
auditor
Set up and run a QUEUE AUDITOR -- a read-only agent, started fresh for each audit, whose one job is to make sure the human's requests are not lost or quietly deprioritized. Its memory lives in files (a charter, a ledger of the human's own words, an unfinished-work ledger, the newest AUDIT file), not in a long thread. It audits at stage boundaries against the queue and the actual work, and writes…
skills/auditor/ on GitHub (opens in a new tab)SKILL.md -
zikr
Recall what the human actually said by querying the thread index (tools/thread_index.py) at four set moments -- before saying "I can't find it", when the human's memory and yours differ, before a decision that may repeat or contradict a past ruling, and right after a context compaction. Cite the turn and date, then re-verify the present before acting. Use whenever a conversation's history matters…
skills/zikr/ on GitHub (opens in a new tab)SKILL.md
No skill matches that filter.
Review before you install
A skill is a set of instructions and scripts your agent will follow, and it can run code on your machine. Read the SKILL.md and any scripts first — this listing is not a security review.