Claude Tools / MCP Servers / ethos-dev
MCP Servers · punt-labs/ethos
ethos-dev
Identity binding for humans and AI agents — voice, email, GitHub, writing style, personality.
CLEARED_STATICmcp-portablev4.13.0
Pipeline results
Coverage
CLEARED_STATIC
Tests run
7 / 8
Portability
MCP-portable
Type
MCP Servers
Tests & results
What the pipeline checked
What's inside
Components found
commands/ext-dev.md
componentmedium
commands/ext.md
componentmedium
commands/identity-dev.md
componentlow
commands/identity.md
componentlow
commands/personality-dev.md
componentmedium
commands/personality.md
componentmedium
commands/role-dev.md
componentlow
commands/role.md
componentlow
commands/session-dev.md
componentlow
commands/session.md
componentlow
commands/talent-dev.md
componentmedium
commands/talent.md
componentmedium
commands/team-dev.md
componentmedium
commands/team.md
componentmedium
commands/writing-style-dev.md
componentmedium
commands/writing-style.md
componentmedium
.claude/agents/ach.md
componentmedium
.claude/agents/adb.md
componentmedium
.claude/agents/adt.md
componentmedium
.claude/agents/ahj.md
componentmedium
Coverage
Not assessed
- behavioural execution not run
- egress patterns logged (FLAG-INFO): 5 lexical match(es)
README
From the repository
# ethos > A responsible agent harness — control, auditability, and > performance for AI agent delegation. [](LICENSE) [](https://github.com/punt-labs/ethos/actions/workflows/test.yml) [](https://pkg.go.dev/github.com/punt-labs/ethos) [](./prfaq.pdf) Ethos binds a name, personality, writing style, domain expertise, email, GitHub handle, and voice into one identity that other tools read, and adds typed mission contracts that record and bind what an agent is asked to do. It ships as a single Go binary with a Claude Code plugin, an MCP server, and a filesystem layout other tools read directly. It runs locally — no server, no telemetry, no cloud. **Platforms:** macOS, Linux (amd64, arm64). ## The Problem Agents write code you're responsible for, and you can't see what they did or why. A developer delegates a task to an AI agent. The agent reads files, edits code, ru…