agentleFS
infeniainfenia/yukta/web/CLAUDE.md

yukta / web

CLAUDE.md1 starsChanged 20 days agoApache-2.0View on GitHub

The file

We split this CLAUDE.md into the rules it gives an agent

1 rule. About 0 are common, 0 are distinctive, and 1 is very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈0

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 1

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    Follow the tiered "Test Execution Policy (speed)" in the root CLAUDE.md: run --tests for the classes you changed first.

    line 35

≈ “Common” is an estimate: at least 1 in 5 of the 27,641 CLAUDE.md files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
CLAUDE.md

Web Module CLAUDE.md

Overview

Web layer with REST controllers and UI support.

Build & Development Commands

# Run web tests
yk gradle :web:test

Patterns

  • Controllers return Mono<ApiResponse<T>> (or a Flux of ServerSentEvent). They never catch errors: core throws NotFoundException, GlobalExceptionHandler maps the status.
  • Execution routes live at /api/executions/{id}/…; the /api/workflow/… paths are deprecated aliases. Use ExecutionAccess for session-ownership checks, Pagination for lists and SseStreams for SSE.
  • GET /api/plugins/{type} includes configSchema; POST /api/sessions/validate dry-runs a draft workflow with WorkflowValidator (always 200, valid=false plus the first problem) and returns each node's output ports; GET /api/sessions/{id}/config returns the full config in the shape POST /api/sessions accepts.
  • /api/supervisor-engines is the engine catalog (/{type}, /{type}/presets) and /api/supervisors the configured supervisors, their /{id}/assignments (GET, PUT, /check; only MANUAL ones are writable, DEFINITION and SYSTEM give 409; 409 OWNERSHIP_CONFLICT names the other owner), /validate, /decisions, /{id}/versions, /rollback, /replay, /{id}/engine/rules (GET/PUT the rules document of a rules supervisor, and POST .../replay a draft; common settings stay, 409 for another engine; the logic is SupervisorRules in core, shared with MCP) (SupervisorsController, SupervisorTypesController). GET /api/supervisors/status lists the running supervisors (SupervisorHost.status(): id, kind, type, dry-run, and applied/rejected/recorded/dropped command counts since startup); the UI shows it on the Control page. /api/agent/v1/** is the agent API of the http engine (AgentApiController, AgentCredentialFilter; a supervisor credential only, never a person's key) and /api/agent/v1/mcp its MCP form (AgentMcpController).
  • yukta.api-keys (optional) guards /api/** and the actuator (except /actuator/health); the server binds to loopback unless YUKTA_SERVER_ADDRESS is set, and refuses to start on another address without a key (ExposureGuard in boot). yukta.api-keys[i].{name,key,role,sessions} define the keys, each with a role: VIEWER (GET only), OPERATOR (also POST/DELETE under /api/executions/ and /api/workflow/), ADMIN (everything, including sessions/workflows, scheduler limits, raw node commands and the actuator). A key is plain text or sha256:<hex>; several keys at once allow rotation. A key with sessions reaches only /api/sessions/{s}/**, /api/workflow/{s}/** and /api/executions/{id}/** (runs in those sessions only). A role or session that does not allow a request gets 403; with named keys the control audit records the actor key:<name>. The path is matched after decoding; ;, encoded / \ . %, // and dot segments give 400 (RequestPaths). HostOriginFilter (only without a key) answers 403 to a Host that is not loopback / yukta.allowed-hosts, and to a non-GET whose Origin is not the Host / yukta.allowed-origins. MCP (McpEndpointFilter) uses the same keys and checks each tools/call against the key: VIEWER reads, OPERATOR also starts and controls, ADMIN does everything (McpToolAccess); a session-limited key needs a sessionId of its own.
  • GET /api/workflow/{s}/history takes workflowId, status, text (in the execution id), since and until (start time, ISO date-time) before it pages (HistoryFilter).
  • GET /api/executions?sessionId=&limit=&workflowId=&status=&text=&since=&until= searches the runs of all sessions, newest first (ExecutionSearchController; each session's history is bounded by yukta.execution-log.retained-runs; a key limited to sessions cannot use it).
  • Input requests (InputRequestsController, backed by InputRequestQueries in core): GET /api/sessions/{s}/inputs?state=&workflowId=&executionId=&limit=&cursor= (VIEWER, newest first, default state=OPEN,ESCALATED, unknown state 400), GET /api/executions/{e}/inputs/{id} (with the answer once answered), POST .../answer ({decision, text?, data?, rationale?, expectedVersion?}) and POST .../escalate ({reason, advice?}) under /api/executions/ so OPERATOR and session keys work as for other run controls. A request of another run is 404; answered is 409 with errors decision and responder; expired or cancelled 409; InputForbiddenException 403 with errors rule; wrong shape 422; InputLimitException 429; standby 503. The legacy /signal and /approvals stay until #1805.
  • GET /api/control/audit?limit=&executionId= lists the newest control commands from the execution log (applied or refused, with the actor); the UI shows it on the Control page.
  • GET /api/events/stream envelopes carry a severity (EventSeverity: ERROR for a failed node/run or an error line, WARN for a rejected command, cancelled node/run or warning line, else INFO); ?severity=WARN sends that and above. ?type= takes a comma-separated list of EventsController.TYPES (case-insensitive); an unknown name is a 400 that lists the valid ones. Each envelope also carries rootExecutionId and depth (the run's place in its tree of sub-workflow runs; a run without a parent is its own root at depth 0). Log lines are rated by keyword (ponytail: in EventSeverity), so a plugin that reports a level should replace it.
  • MapStruct: Used for DTO mapping (AppConfigMapper).

Key Files

  • Controllers: src/main/java/com/infenia/yukta/controller/
  • Mappers: src/main/java/com/infenia/yukta/mapper/
  • Module guide: ../docs/contributing/modules/web.mdx (module guide); open risks, bugs and review findings: GitHub issues labelled web (gh issue list -l web; risks also carry risk); evolution path and quality lenses: ../design/assessments/web.md

Imports

@.claude/rules/coding-standards.md @.claude/rules/git-workflow.md

Testing speed

Follow the tiered "Test Execution Policy (speed)" in the root CLAUDE.md: run --tests for the classes you changed first; full :web:test only when many files changed.

  • GET /api/sessions/{s}/workflows/{w}/versions lists the saved versions of a workflow (newest 20, newest first); POST .../rollback?version=<id> makes one current again (saved as a new version, workflow prepared again).

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.