uipath-for-coding-agents

Mint and refresh UiPath Automation Cloud OAuth tokens via the External Application client-credentials grant. Reads UIPATH_CLIENT_ID and UIPATH_CLIENT_SECRET from .env, requests the scope list from policy.yaml::uipath_scopes, writes the live access token to UIPATH_ACCESS_TOKEN so the uipath CLI can use it without re-auth, and refreshes proactively before the 1-hour expiry. Use this skill at the start of every Conductor run, before any uipath CLI or uipath-python SDK call, and whenever Sentry emits a `kind: auth_failed` event.

Cluster failure events by structural fingerprint and append to memory. Used by Diagnostician and Surgeon. A fingerprint is `<top-level-kind>/<refinement>` plus locality (workflow file or agent name). Clustering uses kNN over a structured-feature embedding stored in SQLite. Output is a cluster ID and confidence score; downstream agents use these to decide auto-fix vs HITL.

Validate `policy.yaml` against `policy.schema.json`, expand env-var references, and run a policy dry-run that simulates how Conductor would behave on the current backlog without dispatching anything. Used at boot, before every Conductor schedule, and on-demand via `aurora policy validate` / `aurora policy dry-run`. Hard fails if the policy is invalid; soft warnings for sub-optimal configurations (low test coverage floor, missing emergency_patch gate, etc.).

Read scoped slices of AURORA's three-tier memory (project, org, skill). Used by every agent at the start of a task to load only the context relevant to its work — never the whole memory blob. Backed by SQLite for fingerprints and plain Markdown for narrative memory; access is gated by scope tags so a Build agent can't accidentally pull Operate-fleet failure traces, etc. Reading is the only public interface; writing is via `aurora-fingerprint` (clusters) or direct file I/O within an agent's own scope.

Nightly self-improvement loop. Reads the day's `learnings/<date>.jsonl`, clusters them by skill/agent, identifies recurring patterns (≥ 3 occurrences across ≥ 2 projects), and opens a GitHub PR against `skills/` (or `agents/`) with a proposed update — e.g., a new fingerprint refinement, an updated SKILL.md guidance, a new helper script. The PR is HITL-gated via `aurora-promote` (kind: skill_compost_pr) and never auto-merged. The mechanism that makes the swarm get smarter with use.

Retire an unused or superseded UiPath process safely. Used by Strategist + Auditor + Concierge in sequence: Strategist proposes, Auditor checks dependencies, Concierge gates via Action Center, then this skill executes the deprecation steps. Stops scheduled triggers, archives the package, transfers any in-flight queue items to a successor (if any), updates org memory's deprecation log, and writes a runbook so the retired process can be restored within 30 days if needed.

Extract structured friction signals from raw text sources (Slack messages, Jira issues, email digests, calendar invite notes, screen-recording transcripts). Used by `scout` to convert noisy human chatter into machine-actionable candidate signals. Returns one JSON object per signal with frequency hints, duration hints, pain indicators, and named systems. Reject one-off, internal-swarm, and metadata-only chatter.

Author a Process Definition Document with calibrated ambiguity scoring. Used by `analyst` after a candidate is promoted from Curator's backlog. Templates and the ambiguity rubric live in `templates/`. Output is a structured PDD that Architect, Forgers, and Tester all consume verbatim. Includes a Given/When/Then acceptance-criteria pattern that maps 1:1 to test cases. Sets ambiguity_score in [0,1]; values above 0.4 trigger Interviewer.

HITL gate via UiPath Action Center. Creates a Form Task in the configured catalog (`aurora_supply_chain_approvals`), assigns approvers from a configured list, polls for completion, applies the policy.yaml gate's timeout behavior, and returns the approver's response to the calling agent. Used for prod publishes, emergency patches, deprecations, large fixes, and skill-compost PR reviews. The implementation of every gate in `policy.yaml::gates`.

Replay a faulted Maestro instance (or standalone job) in a sandbox folder using the same input data, to isolate whether the root cause is in the bot, the environment, or an external dependency. Used by Diagnostician for non-trivial faults (cluster overlap < 0.6 or external dependency suspected). The "Shadow twin" pattern from AURORA's design.