AugmentClaude

CORAL Debug

Debug and verify changes to the CORAL framework with targeted reproduce loops.

Installation

  1. Make sure Claude is on your device and in your terminal.

    Skills load from ~/.claude/skills/ when Claude Code starts up β€” so you need it on your machine first. If you don't have it yet, install it once with the command below, then run claude in any terminal to verify.

    One-time setup
    npm i -g @anthropic-ai/claude-code

    Already have it? Skip ahead.

  2. Paste into Claude Code or into your terminal.

    This copies the whole skill folder into ~/.claude/skills/coral-debug-human-agent-society/ β€” the SKILL.md plus any scripts, reference docs, or templates the skill ships with. Safe default: works for every skill.

    Faster alternative (instruction-only skills)

    Skips the clone and grabs only the SKILL.md file. Don't use this if the skill ships Python scripts, reference markdowns, or asset templates β€” they won't be downloaded and the skill will fail when it tries to load them.

    Quick install (SKILL.md only)
    Sign up to copy
  3. Restart Claude Code.

    Quit and reopen Claude Code (or any other agent that loads from ~/.claude/skills/). New skills are picked up on startup.

  4. Just ask Claude.

    Skills auto-activate when your request matches the skill's description β€” no slash command needed. Trigger phrases live in the skill's own frontmatter; you can read them in the β€œWhat this skill does” section above.

Prefer to read the source first? Open on GitHub.

When Claude uses it

Verify and debug changes to CORAL itself β€” smallest reproduce loop per area (grader / daemon / CLI / hooks / manager / workspace / hub / template / config / web), where to look when something breaks (hung graders, agent restart loops, stalled agents, missing heartbeat actions, corrupted shared state, broken worktree symlinks, grader import errors, wrong-task resume), how to inspect a live or finished run under `.coral/public/`, and the canonical lint/test commands. Use when editing code under `coral/` or chasing a CORAL bug, NOT when adding a new task or extending the framework.

What this skill does

CORAL debug & change-verification workflows

This skill is for people (and Claude Code) hacking on the CORAL package itself, not for users authoring a task. For authoring guides see the siblings:

  • coral-new-task β€” creating a new examples/<task>/ (seed + task.yaml + grader)
  • coral-extend β€” extending CORAL itself (new runtime, new CLI command, new bundled skill, ...)

Reproduce loops

Pick the smallest one that exercises your change.

Editing...Fastest reproduce
coral/grader/{task_grader,loader,subprocess_grader}.py or a builtin graderuv run coral validate examples/circle_packing (or any example with a packaged grader)
coral/grader/daemon.py (parallel grading, queueing, worktree isolation)uv run pytest tests/test_grader_daemon.py -v
coral/grader/{base,protocol}.pyuv run pytest tests/test_grader.py tests/test_subprocess_grader.py -v
coral/cli/*.py (argparse, dispatch, output formatting)Run the command directly: uv run coral log --help, uv run coral status, etc. against a latest run under examples/<task>/results/
coral/hooks/post_commit.py (submit_eval)uv run pytest tests/test_hooks.py -v, then end-to-end via coral start -c task.yaml agents.count=1 and watch .coral/public/attempts/
coral/agent/manager.py, state.py, exit_classifier.pyuv run pytest tests/test_manager_reliability.py tests/test_manager_seen_attempts.py -v
coral/agent/builtin/*.py (a runtime)uv run pytest tests/test_<runtime>.py if it exists, then coral start -c task.yaml agents.runtime=<name> agents.count=1
coral/workspace/{project,worktree,repo,grader_env}.pyuv run pytest tests/test_workspace.py tests/test_grader_env.py -v
coral/hub/{attempts,notes,skills,checkpoint,heartbeat}.pyuv run pytest tests/test_hub.py tests/test_checkpoint.py tests/test_heartbeat.py -v
coral/template/coral_md.py or templatesuv run pytest tests/test_template.py -v
coral/config.pyuv run pytest tests/test_config.py -v
coral/web/* (dashboard)uv run coral ui against any existing run; reload the browser after edits

End-to-end smoke (slow but exercises the whole pipeline):

uv run coral start -c examples/circle_packing/task.yaml agents.count=1 run.session=local
# Wait for one eval to land in .coral/public/attempts/, then:
uv run coral stop

Where to look when X breaks

SymptomFirst place to look
Grader hangs / pending attempts pile up.coral/public/grader_daemon.pid (is daemon alive?), .coral/public/grader_daemon_heartbeat (mtime), .coral/public/eval_logs/<hash>/ for grader stdout/stderr
coral eval errors outThe pending JSON itself: .coral/public/attempts/<hash>.json β€” status and feedback fields. Source: coral/hooks/post_commit.py::submit_eval
Agent restart loopcoral status shows pause state. Source: coral/agent/manager.py (restart_burst_threshold, restart_burst_window). Per-agent runtime stdout/stderr is captured under the worktree.
Stalled agentagents.timeout watchdog in coral/agent/manager.py. Grader-queue exemption (grader_pending_max_age) skips the watchdog while a recent submission is still pending.
Heartbeat actions not firing.coral/public/heartbeat/<agent_id>.json (per-agent action list), .coral/public/eval_count (global counter). Logic: coral/agent/heartbeat.py::HeartbeatRunner.check.
Shared state corrupted / race.coral/public/.git/ is a real git repo β€” git -C .coral/public log shows checkpoint history; coral notes --history is the friendly view.
Worktree symlinks missingcoral/workspace/worktree.py β€” symlinks .coral/public/ into each agent worktree under the runtime-specific name (.claude / .codex / .opencode).
Grader can't import its packageCheck .coral/private/grader_venv/ β€” setup_grader_env ran uv venv + grader.setup once at run start. Re-running coral start on the same run_dir does NOT re-bootstrap; delete the venv to force it.
Resume picks up wrong taskcoral resume reads .coral/config.yaml and .coral/config_dir. The latest symlink at results/<task-slug>/latest decides which run is resumed.

Inspect a live or finished run

RUN=$(readlink -f results/<task-slug>/latest)
ls "$RUN/.coral/public/"            # attempts/, notes/, skills/, agents/, logs/, eval_logs/, eval_count
cat "$RUN/.coral/public/attempts/<hash>.json" | jq .
ls "$RUN/agents/"                    # one worktree per agent
cat "$RUN/.coral/config.yaml"        # exact config used (post-merge)

The web dashboard (coral ui) reads the same files; if you're debugging dashboard rendering, coral/web/api.py is the route handler and coral/web/static/ is the built React bundle.

Test + lint

uv sync --extra dev                  # one-time, gets pytest + ruff
uv run pytest tests/ -v              # full suite (~seconds, no docker)
uv run pytest tests/test_<thing>.py -v -k <pattern>
uv run ruff check .                  # lint
uv run ruff format .                 # autoformat

There is no separate type-check step in CI yet; tests cover the contract.

Quick task scaffold for ad-hoc testing

When you need a throwaway task to exercise a code change:

uv run coral init /tmp/coral-scratch
# edit /tmp/coral-scratch/grader/src/coral_scratch_grader/grader.py to return a simple float
uv run coral validate /tmp/coral-scratch
uv run coral start -c /tmp/coral-scratch/task.yaml agents.count=1 run.session=local

coral validate runs the grader against seed/ in a tempdir without spawning agents β€” the fastest way to confirm a grader change compiles and produces a Score.

Conventions

  • No git from agents. All commits go through coral eval. If you're adding a feature that needs to commit, route it through coral/hooks/post_commit.py or coral/hub/checkpoint.py.
  • Atomic writes for any JSON in .coral/public/. Use coral.hub.attempts.write_attempt (tmp + rename) as the model.
  • Idempotent setup. setup_run_dir, setup_grader_env, init_checkpoint_repo must all be safe to call twice.
  • Shared dir is configurable. Don't hardcode .claude; thread shared_dir through (see coral/template/coral_md.py).
  • Lazy imports in CLI dispatch (coral/cli/__init__.py) β€” keep coral --help fast.

Related skills