API Overview

The Hermes dashboard API is the handoff point between three different actors:

• browser clients used by the dashboard frontend
• Hermes agents that fetch work, save outputs, and review feedback
• scraper services that ingest new job descriptions

Use this reference to understand which endpoints belong to which part of the workflow and what each call is expected to do.

Scope of this reference

These pages document the endpoints currently used by this repository and its documented Hermes workflows.

You are not required to copy this API surface exactly. If your setup needs additional routes, different route names, or extra workflow-specific endpoints, you can design them that way as long as the skills and frontend code that depend on them are updated accordingly.

Treat this section as the current working contract for this repo, not as a rule that your backend can never extend.

Base URL

All examples in this section use:

<DASHBOARD_BASE_URL>

Keep the real deployment URL out of shared docs, prompts, and examples.

Authentication modes

• Website session auth: used by browser-based dashboard features such as login, logout, and current-user lookups.
• API key auth: used by Hermes agents and other trusted server-side clients with an Authorization: Bearer <API_KEY> header.
• Scraper ingest auth: used by scraper services that push job descriptions into the queue with a dedicated shared secret or ingest token.

Security rules

• Do not send user_id from frontend clients unless an endpoint explicitly requires it.
• Ownership is resolved server-side from the authenticated user or API key.
• Hashes, internal secrets, and other credential material should never appear in responses.
• Use placeholders such as <API_KEY>, <SCRAPER_INGEST_SECRET>, and <JOB_DESCRIPTION_ID> in examples and setup docs.

How to read this reference

• If you are building the dashboard UI, focus on Authentication APIs, Feedback APIs, and Hermes Integration APIs.
• If you are wiring the Hermes orchestrator, focus on Job Description APIs, Resume Output APIs, Workflow Logging APIs, and Feedback APIs.
• If you are building a scraper, focus on POST /api/job-descriptions plus the scraper-specific authentication notes.

Current skill usage map

This map shows where the currently documented endpoints are used inside the Hermes workflow.

Endpoint	Current skill usage	Workflow part
GET /api/job-descriptions/next	resume-pipeline-orchestrator	Batch intake at the start of the pipeline
POST /api/generated-resumes	resume-pipeline-orchestrator	Final artifact persistence after LaTeX assembly and self-review
PATCH /api/job-descriptions/:id/use	resume-pipeline-orchestrator	Queue state update after a successful resume push
POST /api/workflow-logs	resume-pipeline-orchestrator	Per-JD status logging, orchestrator error logging, and batch summary logging
GET /api/feedback/pending	feedback-processor related workflow	Fetch pending human feedback for review
PATCH /api/feedback/:id/review	feedback-processor related workflow	Write Hermes review output back to the dashboard
POST /api/feedback	Dashboard frontend	User submits feedback on generated output
POST /api/integrations/hermes	Dashboard frontend	Save Hermes connection/configuration
POST /api/hermes/test	Dashboard frontend	Verify the saved Hermes connection
POST /api/hermes/dispatch	Dashboard frontend	Trigger Hermes work from the dashboard
POST /api/auth/login	Dashboard frontend	Start session
POST /api/auth/logout	Dashboard frontend	End session
GET /api/auth/me	Dashboard frontend	Resolve current session user
POST /api/auth/register	Dashboard frontend	Account creation
POST /api/auth/api-key	Dashboard frontend and resume-pipeline-orchestrator setup	Issue the API key later used by Hermes

Reference sections

• Job Description APIs
• Resume Output APIs
• Workflow Logging APIs
• Feedback APIs
• Hermes Integration APIs
• Authentication APIs

If you want the actual skills and repository workflows that consume these endpoints, use the public source repository.

hermes-autonomous-resume on GitHub