opendatadiscovery
OrganizationAI maintainer team for Open Data Discovery — coordinates audit, gap-closing, and sustainable maintenance across ODD repositories using navigation indexes, coverage-tracked scanners, and structured work items.
Categories
Indexed Skills (18)
contribute
Resolve a GitHub issue on opendatadiscovery/odd-platform end-to-end as a virtual contributor — scope, reproduce, root-cause, plan (GATE 1), change code, add unit + integration tests, update docs + ontology, open a DRAFT PR (GATE 2). Composes the ontology / test / doc / adr machinery; posts clarifying + root-cause comments via a scoped-token GitHub App; never merges.
coverage
Show scan coverage across all scanners — what's been analyzed, what remains, what changed since last scan.
enrich
Run the file-analyser subagent against one or more substrate nodes to produce per-node semantic enrichment sidecars. Resolves a path / node-id / batch spec to substrate nodes; spawns file-analyser per node; validates the resulting sidecar against the schema; updates the manifest's `enrichment` block.
enumerate
Generate or refresh coverage manifest for a scanner — lists ALL items in scope so scans can track progress across sessions.
implement
Work a batch of backlog items end-to-end as a maintainer — hold the Implementation Quality Bar, run the Pre-authoring stance check, log follow-ups on disk, flip items to `review-ready`, and ship one PR per repo per batch.
next-batch
Autonomous batch driver for the rev-2 agentic-ontology sprint. Picks the next pending theme from state/sprint-themes.yaml, spawns 5 file-analyser subagents in parallel, then 5 reducers in parallel, merges deltas, refreshes coverage, appends the investigator-log entry, commits, pushes to remote, marks the theme done. Designed to be fired by `/loop` overnight without maintainer input.
panel
Run the methodology meta-review — a single methodology-reviewer agent that traces the whole current methodology (APPROACH.md, the ADRs, the agent contracts, the skills, the playbooks, the case-law, the live artefacts), diffs against the prior review, runs fresh blind spot-checks against real source, and emits real gaps + improvement proposals at lineage/{repo}/meta-reviews/{date}/review.md. `/panel` runs it; `/panel --show` prints the latest; `/panel validate` runs the acceptance-corpus check.
probe-define
Generate probe-skeleton YAML files at `lineage/{repo}/probes/{P-NNN}.yaml` from `feature-flows.yaml`'s uncovered test-matrix cells. Each skeleton carries the feature_id + test_class + cross-references + a TODO-marked arrange/act/observe/assert block the maintainer fills in. Accelerates probe corpus growth — slice 4 of dynamic-verification ADR. Local-only per APPROACH.md section 5 rule 12.
probe-run
Run one or more dynamic-verification probes against a local ephemeral docker-compose mirror of the target system. Captures observation traces as committed artefacts at `lineage/{repo}/probe-runs/{date}-{probe_id}.yaml` and merges measured values back into `feature-flows.yaml`, `test-map.yaml`, and per-sidecar `confidence_per_field` annotations. Local-only execution per APPROACH.md section 5 rule 12 + section 9 — no remote infrastructure ever. Slice 2 of `adrs/drafts/dynamic-verification-layer.md`.
probe
Run the probe-driven validation rounds for the agentic ontology — Type 4 (adversarial — capability-negation / cross-product-fabrication / synonym-swap-with-negation, ≥2 of 3 PASS) and Type 6 (implicit-ADR confirmation — maintainer writes 5 ADRs they know are followed; ≥3 of 5 must surface in implicit-adrs.md top-10). Both rounds are MAINTAINER-AUTHORED; this skill is the protocol mechanics, not the answers.
reflect-feature
Run the feature-reflector subagent (Layer 4b — top-down product-owner reflection) over one composed feature flow. Produces `lineage/{repo}/feature-reflections/detail/{F-NNN}.yaml` with a stepped-back product-owner narrative + 5-15 falsifiable user-facing hypotheses + per-hypothesis verdicts (confirmed / contradicted / partial / probe-needed) traced through the implementation chain. Contradictions are surfaced as bug-candidate / caveat-candidate findings — cross-file intent-vs-implementation drifts no single sidecar can see in isolation. Per APPROACH.md section 15 + rev-5 LSN-020 case-law.
review
Verify a `review-ready` work item (or an entire batch) meets every acceptance criterion and every Quality Bar gate. Reject by default — each check requires cited evidence. Must run in a session distinct from the `/implement` session that produced the item.
scan
Run an audit scanner against ODD repositories. Reads scanner definition, picks unscanned items from coverage manifest, writes findings, updates navigation.
triage
Convert raw scanner findings into atomic work items in the backlog. Assigns priority, category, effort, and identifies file conflicts.
ingest-docs
Ingest the live documentation manual into the ontology as ground-truth Doc nodes, then enrich them with doc→ontology DESCRIBES links. Runs the mechanical doc extractor (docs-ingest → doc-nodes.jsonl, split by heading/anchor, prose referenced from ../documentation), rebuilds the graph (embeds the doc prose), spawns the doc-analyser subagent per un-enriched page (live-URL verification + DESCRIBES to the concepts/features/code each page documents + doc-claim-vs-code drift), and reports the consistency dashboard. Per adrs/drafts/ground-truth-lineage.md (Phase 1).
navigate
Find where a feature, component, or concept lives in the ODD codebase. Uses the navigation index for O(1) lookups.
orient
Quick orientation for a new session — shows system overview, current state, and available actions.
status
Show current progress of the audit and implementation system — scanner completion, backlog counts, blockers.
Bio shown is the top-scored skill's repo description as a fallback — real GitHub bios land in a future update.