jmagly/aiwg

Agent Loop detects requests for iterative autonomous task execution and routes them to the right loop executor. Use it when a user wants to keep trying, fix until green, or run an explicit agent loop.

Agent Loop executes a task in repeated cycles until verifiable completion criteria are met. Use it when you want an agent to keep trying, learn from failures, and report the final status.

Agent Loop External launches a long-running agent loop outside the current session with saved state. Use it for crash recovery, parallel runs, and CI/CD-backed tasks that need resumable execution.

Agent Persistence defines reusable human-in-the-loop gates for persistent agent loops. Use it when you need authorization for destructive actions, policy overrides, or recovery escalation.

Al External launches long-running iterative loops outside the AI session with saved state. Use it for tasks that need crash recovery, parallel runs, or CI/CD integration.

Al Memory lets you list, query, and clear semantic memory entries from loop iterations. Use it to review lessons learned, failure patterns, and success patterns across runs.

Auto Test Execution runs relevant tests after code-generating agents modify source files. It is used to enforce an execute-before-return workflow and retry when tests fail.

Cockpit Headless Deployment guides safe deployment or repair of Cockpit and Agentic Sandbox on a headless Linux host. Use it when host topology, SSH access, persistence, or runtime readiness must be planned and verified before changes.

Cross-Task Learner extracts reusable patterns from completed agent loops and injects relevant ones into new loops. It helps agents reuse fixes, success paths, anti-patterns, and code templates across tasks.

Debug Memory queries and manages executable feedback session history. Use it to find past failures, inspect learned patterns, review stats, or archive the memory store.

Execute Feedback runs tests on generated code, analyzes failures, applies targeted fixes, and retries until the code passes or max attempts are reached. Use it after code generation or edits to verify behavior.

Flow Syntax validates YAML flow syntax against the pinned reference parser. Use it when you need to check whether input is accepted by that parser.

Infer Completion Criteria derives a measurable pass/fail gate for an agent-loop task when the user did not supply --completion. It reads project docs, manifests, CI, and AIWG artifacts to pick an unambiguous completion check.

Minimal validates the minimal normative Agent Skills fixture. Use it when checking that a validation workflow or cataloging pipeline handles the simplest possible skill entry.

Mission Control orchestrates multi-loop background work from a dashboard. Use it to start sessions, dispatch missions, monitor progress, and stop or pause running work.

Portable Complete uses a complete portable fixture to verify Agent Skills round trips. It points you to a guide, asset, license, and verification script to inspect without running it during import.

Ralph Abort stops a running or paused agent loop. Use it to keep the current changes or revert the loop’s commits and clean up loop state.

Ralph Analytics shows aggregate metrics from agent loop execution history. It is used to review success rates, iteration counts, stuck loops, reflections, and escalation patterns.

Ralph Attach attaches to a running external agent loop and streams its live output log. Use it to watch progress or detach with Ctrl+C without stopping the loop.

Ralph Config shows and changes agent loop settings like iteration limits, timeouts, provider, and commit behavior. Use it to view, set, reset, or apply presets that persist to .aiwg/ralph/config.json.

Ralph Reflect lets you view, search, and clear agent loop reflections and episodic memory. Use it to inspect outcomes, recurring patterns, and strategy changes across loops.

Ralph Resume continues an interrupted agent loop from the last checkpoint. Use it when a previous run stopped early due to limits, timeout, or interruption.

Ralph Status checks the current or latest agent loop state from local .aiwg files. Use it to view progress, iteration history, completion reports, or errors when a loop is active or missing.

Reflection Injection loads relevant past reflections into an agent's context at the start of an iteration or retry. Use it to carry forward lessons from prior failures and avoid repeating mistakes.

Setup Generate creates a setup.manifest.yaml for a project using the setup.aiwg.io/v1 language. Use it when you need a starter install workflow for user, developer, or CI setup.