Skip to content

levi-qiao/longgraph-skill / skills/loop-graph

loop-graph

Flagged

Use only when the user explicitly invokes /loop-graph or names this skill. Compiles a custom loop-graph run. Do not use unless named. Existing runtime nodes are self-contained.

references/shell-cron.md

View on GitHub
# Shell / cron

Read only for nodes hosted by an agent CLI under shell or cron.

## Runtime shape

- Each CLI process starts fresh, so this is the reference fresh-per-tick supervisor.
- Prefer a sequential loop: run one round to closure, inspect ledger terminal state,
  then sleep. Use cron only when its lock and job-removal commands are already known.
- Stop with `break` or remove the cron job at terminal state.
- Delete `TIMER_STEP` and the `ops.md` Timers section; each process is already fresh and
  the wrapper owns scheduling and stop. Never ask the owner to write a process or job
  ID into `ops.md`.

## Fill the generic handoff

- `SESSION_INSTRUCTION`: Run executor and supervisor as separate processes against
  the same workspace.
- `EXECUTOR_DESTINATION`: executor process.
- `EXECUTOR_LAUNCH`:

  ```sh
  while :; do
    {{EXECUTOR_CLI_COMMAND}} "Read {{RUN_DIR}}/executor.md and follow it exactly"
    grep -qE '^Run status: `(exit-ready|stalled|closed)`' {{RUN_DIR}}/ledger.md && break
    sleep {{EXEC_INTERVAL}}
  done
  ```

  Resolve `EXECUTOR_CLI_COMMAND` from the workspace before delivery. If no concrete
  command is available, choose another host; never print `<cli>` or ask the owner to
  invent the invocation.

- `EXECUTOR_READY`: the first process starts and acquires any configured lock.
- `SUPERVISOR_DESTINATION`: separate cron entry or sequential process.
- `SUPERVISOR_LAUNCH`: provide the same concrete sequential wrapper for
  `{{SUPERVISOR_CLI_COMMAND}}`, with the anchored Run-status check above. Never print
  an unresolved command. Use cron only when the exact lock and removal commands can
  also be filled.
- `RUNNING_STATE`: both processes use only the run directory for shared state.
- `RESET_INSTRUCTION`: none; each CLI process already starts fresh.
- `STOP_INSTRUCTION`: break both loops or delete their cron entries.
reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis.Stars: 469kUpdated Claude CodeCodexVerifiedInstall
diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness.Stars: 321kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 321kUpdated Claude CodeCodexVerifiedInstall
executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is availableStars: 321kUpdated Claude CodeCodexVerifiedInstall