epic-harness

devops-infra MCP Server

Multi-tool AI agent harness — 22 skills, self-evolving engine, unified memory, autonomous spec-to-PR pipeline. Works with Claude Code, Codex, Cursor, OpenCode, and Cline.

Verified
devops-infradevops-infra
5 views8 stars1 forksv0.6.5Apache-2.0

Why This Matters

Discovered via agent-topic:ai-agent and last synced 3mo ago.

Verified
Source
agent-topic:ai-agent
Stars
8
Last synced
3mo ago
Install
Check source

Install

Install instructions not detected yet

Check the source repository for the latest setup steps.

View source instructions
44
Tools
0
Resources
0
Prompts
Standard I/O
Transport

Available Tools (44)

Technique

How it works

repeated_same_error

Same error N+ times

Go

`evo-go-care`

Stream-aligned

`stream`

resolution

Manual / MCP

fix_then_break

Edit success → build/test fails

Python

`evo-py-care`

Skill

With

Hook

When

subsystem

domain-specialist, integration-tester

instinct

Auto (reflect)

FAIL

RETRY{"retry < 3?"} RETRY -->

Profile

Active hooks

Tool

Purpose

pattern

Auto (reflect)

no

PAUSE["Pause\nuser decides"]:::human PAUSE -->

evo-ts-care

0.87

minimal

guard, observe, resume

Command

What it does

error

Auto (reflect)

long_debug_loop

Stuck on same file

Rust

`evo-rs-care`

Enabling

`enabling`

project

Manual / MCP

evo-bash-discipline

0.65

Pattern

Detects

Stack

Presets

Type

Created by

decision

Manual / MCP

Issue

When

strict

all hooks + future strict-only checks

session

Auto (reflect)

Platform

`platform`

concept

Manual / MCP

Choice

Why not enhanced

thrashing

Edit↔Error alternating

mem_add

Add node with auto-importance by type (or explicit 0.0–1.0)

abort

ABORT(["Abort"]) SHIP --> EVOLVE["Evolve\nauto-analyze session"]:::auto EVOLVE --> DONE(["Orbit Complete\nconsolidated report"]):::auto classDef human fill:#4a4a6a,stroke:#9b9bcc,color:#fff classDef auto fill:#1a5c3a,stroke:#4caf7d,color:#fff ``` **Purple** — human steps: mode selection (unclear → interactive), 3× check failure pause. **Green** — clear + complex → council auto-spec; clear + simple → direct build; both fully autonomous. State persisted in `$HARNESS_DIR/orbit/PIPELINE-{timestamp}.json` — survives context compaction. > **Caveats**: The agent may bypass the pipeline when modifying orbit itself or editing docs only. See [Known Issues (Agent Judgment)](#known-issues-agent-judgment). --- ## Auto Skills (Ring 2) Skills trigger automatically based on context. You don't invoke them.

mem_context

Project-scoped smart recall (no hint)

Skills

Agents

mem_related

Graph traversal from a node ID (finds connected knowledge)

mem_search

Keyword search (full-text), ranked by importance

mem_recall

Smart contextual recall with hint + project + graph neighbors

mem_query

Filter by tag/type/project