Skip to content

Refactor runtime to use pluggable agent harnesses #2

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Open
andrewhampton wants to merge 1 commit into
base: master
Choose a base branch
from
Open

Refactor runtime to use pluggable agent harnesses #2

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
24 changes: 18 additions & 6 deletions README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
@@ -1,6 +1,6 @@
# Cord

A coordination protocol for trees of Claude Code agents.
A coordination protocol for trees of agent workers.

One goal in, multiple agents out. They decompose, parallelize, wait on dependencies, and synthesize — all through a shared SQLite database.

Expand Down Expand Up @@ -31,8 +31,9 @@ No workflow was hardcoded. The agent built this tree at runtime.
## Prerequisites

- [uv](https://docs.astral.sh/uv/) (Python package manager)
- [Claude Code CLI](https://docs.anthropic.com/en/docs/claude-code) installed and authenticated (`claude` command available)
- An Anthropic API key with sufficient credits
- At least one supported agent CLI installed and authenticated:
- [Amp CLI](https://ampcode.com/) (`amp` command available)
- [Claude Code CLI](https://docs.anthropic.com/en/docs/claude-code) (`claude` command available)

## Install

Expand All @@ -51,7 +52,13 @@ cord run "Analyze the pros and cons of Rust vs Go for CLI tools"
# Or point it at a planning doc
cord run plan.md

# Control budget and model
# Force a backend
cord run "goal" --amp
cord run "goal" --claude
# (also supported)
cord --amp run "goal"

# Control budget and model (Claude backend)
cord run "goal" --budget 5.0 --model opus
```

Expand All @@ -61,6 +68,10 @@ cord run "goal" --budget 5.0 --model opus
|------|---------|-------------|
| `--budget` | 2.0 | Max USD per agent subprocess |
| `--model` | sonnet | Claude model (sonnet, opus, haiku) |
| `--amp` | auto | Force Amp harness |
| `--claude` | auto | Force Claude harness |

If neither backend flag is passed, Cord auto-detects available CLIs and chooses in this order: `amp` first, then `claude`. If neither binary is installed, the CLI exits with an actionable install error.

## How it works

Expand Down Expand Up @@ -128,7 +139,8 @@ src/cord/
prompts.py # Prompt assembly for agents
runtime/
engine.py # Main loop, TUI
dispatcher.py # Launch claude CLI processes
harness/ # Agent harness abstraction + implementations
dispatcher.py # Compatibility wrapper (Claude launch)
process_manager.py # Track subprocesses
mcp/
server.py # MCP tools (one server per agent)
Expand Down Expand Up @@ -158,7 +170,7 @@ Each agent subprocess has its own Claude API budget (set via `--budget`). A simp
- No web UI — terminal TUI only.
- No mid-execution message injection (pause/modify/resume requires relaunch).
- Each agent gets its own MCP server process (~200ms startup overhead).
- Claude Code CLI must be installed and authenticated.
- Amp CLI or Claude Code CLI must be installed and authenticated.

## Alternative implementations

Expand Down
Loading