agent-loop-ext
Crash-resilient external agent loop with state persistence and CI/CD integration
Plan and delegate safe dataset retirement, tombstoning, derived-artifact cleanup, and reconciliation.
$ npx -y skills add jmagly/aiwg --skill dataset-retire --agent claude-codeHow it fires
How this skill gets triggered: by you, by Claude, or both.
/dataset-retireContext preview
The summary Claude sees to decide when to auto-load this skill.
Plan and delegate safe dataset retirement, tombstoning, derived-artifact cleanup, and reconciliation.
namespace: aiwg name: dataset-retire description: Plan and delegate safe dataset retirement, tombstoning, derived-artifact cleanup, and reconciliation. version: 1.0.0 platforms: [all] aliases: [remove dataset, decommission index, dataset cleanup] triggers: [retire this dataset, remove this index safely, delete derived dataset artifacts, reconcile dataset deletion, decommission this source]
First encode the retirement operation and complete enumeration in a governed plan input, then use `aiwg dataset plan --file <plan-input.json> --json`. Require complete bounded enumeration of affected canonical and derived artifacts, retention/legal-hold checks, tombstones by default, rollback limits, and a reviewed bulk threshold. Dataset Policy Reviewer approval must match the exact plan digest.
Delegate only through `aiwg dataset ingest` with the exact retirement plan digest, idempotency key, approval IDs, preview digest, and threshold. Never delete directly. Preserve canonical material unless explicitly and separately authorized. Finish with `aiwg dataset verify <run-ref> --json` and a reconciliation/evidence handoff.
Reusable project context and specialist workflows for the AI tools you already use. Plan software, coordinate specialist reviews, prepare campaigns, investigate incidents, organize research, curate media, and maintain operational knowledge.
Repo: jmagly/aiwg
Crash-resilient external agent loop with state persistence and CI/CD integration
Detect requests for iterative autonomous agent loops and route to the appropriate loop executor
Automatically execute tests when code-generating agents modify source files, enforcing the execute-before-return pattern
Enable agent loops to learn from similar past tasks and share patterns across loops
Query and manage the executable feedback debug memory
Execute tests on generated code and iterate until passing