OpenAI-compatible REST API for external tool integration:
Chat Completions: POST /v1/chat/completions — stateless, with X-Hermes-Session-Id for continuity.
Responses API: POST /v1/responses — stateful via previous_response_id.
Sessions: CRUD + fork for session management.
Runs: POST /v1/runs (returns run_id immediately), GET /v1/runs/{id} (status), GET /v1/runs/{id}/events (SSE stream), POST /v1/runs/{id}/approval and /stop (resolve or interrupt).
Models: GET /v1/models — list configured models.
Health: GET /health (simple), GET /health/detailed (authenticated readiness).
This enables Open WebUI, LobeChat, and other OpenAI-compatible frontends to use archie-core as a backend.
curl -X POST /v1/chat/completions with a model and messages returns a streaming response. Sessions persist across requests via session ID. Runs can be stopped mid-execution.
Notes
docs/prds/hermes-gap-analysis.md (referenced above) was intentionally deleted in commit 65aa5b4 ('chore: cleanup', 2026-07-28) after it had already spawned this and ~14 sibling beads issues from its 60 identified gaps — it is not a broken link to restore. This issue's description is the surviving record of its slice of that analysis.
Close reason
No use case (maintainer 2026-09-22). Recorded for the file: the premise was NOT invented by the bead author - the deleted source doc claims it twice (hermes-gap-analysis.md:25 'API server | OpenAI-compatible REST | None' and :153 'OpenAI-compatible API server (/v1/chat/completions, /v1/responses, session CRUD, runs with SSE streaming)'). That doc was an AI-generated vendor comparison and the Hermes claim was never independently verified, so this closes as not-wanted rather than as disproven. Do not resurface on parity grounds.
Imported from beads archie-core-1786637496982-273-2078b865 · GitHub #294 · type feature · priority P2 · created 2026-08-08 · closed 2026-09-21
OpenAI-compatible REST API for external tool integration:
1. **Chat Completions**: POST /v1/chat/completions — stateless, with X-Hermes-Session-Id for continuity.
2. **Responses API**: POST /v1/responses — stateful via previous_response_id.
3. **Sessions**: CRUD + fork for session management.
4. **Runs**: POST /v1/runs (returns run_id immediately), GET /v1/runs/{id} (status), GET /v1/runs/{id}/events (SSE stream), POST /v1/runs/{id}/approval and /stop (resolve or interrupt).
5. **Models**: GET /v1/models — list configured models.
6. **Health**: GET /health (simple), GET /health/detailed (authenticated readiness).
This enables Open WebUI, LobeChat, and other OpenAI-compatible frontends to use archie-core as a backend.
Ref: docs/prds/hermes-gap-analysis.md (api_server.py)
### Acceptance Criteria
curl -X POST /v1/chat/completions with a model and messages returns a streaming response. Sessions persist across requests via session ID. Runs can be stopped mid-execution.
### Notes
docs/prds/hermes-gap-analysis.md (referenced above) was intentionally deleted in commit 65aa5b4 ('chore: cleanup', 2026-07-28) after it had already spawned this and ~14 sibling beads issues from its 60 identified gaps — it is not a broken link to restore. This issue's description is the surviving record of its slice of that analysis.
### Close reason
No use case (maintainer 2026-09-22). Recorded for the file: the premise was NOT invented by the bead author - the deleted source doc claims it twice (hermes-gap-analysis.md:25 'API server | OpenAI-compatible REST | None' and :153 'OpenAI-compatible API server (/v1/chat/completions, /v1/responses, session CRUD, runs with SSE streaming)'). That doc was an AI-generated vendor comparison and the Hermes claim was never independently verified, so this closes as not-wanted rather than as disproven. Do not resurface on parity grounds.
---
_Imported from beads `archie-core-1786637496982-273-2078b865` · GitHub #294 · type feature · priority P2 · created 2026-08-08 · closed 2026-09-21_
Blocking a user prevents them from interacting with repositories, such as opening or commenting on pull requests or issues. Learn more about blocking a user.
OpenAI-compatible REST API for external tool integration:
This enables Open WebUI, LobeChat, and other OpenAI-compatible frontends to use archie-core as a backend.
Ref: docs/prds/hermes-gap-analysis.md (api_server.py)
Acceptance Criteria
curl -X POST /v1/chat/completions with a model and messages returns a streaming response. Sessions persist across requests via session ID. Runs can be stopped mid-execution.
Notes
docs/prds/hermes-gap-analysis.md (referenced above) was intentionally deleted in commit
65aa5b4('chore: cleanup', 2026-07-28) after it had already spawned this and ~14 sibling beads issues from its 60 identified gaps — it is not a broken link to restore. This issue's description is the surviving record of its slice of that analysis.Close reason
No use case (maintainer 2026-09-22). Recorded for the file: the premise was NOT invented by the bead author - the deleted source doc claims it twice (hermes-gap-analysis.md:25 'API server | OpenAI-compatible REST | None' and :153 'OpenAI-compatible API server (/v1/chat/completions, /v1/responses, session CRUD, runs with SSE streaming)'). That doc was an AI-generated vendor comparison and the Hermes claim was never independently verified, so this closes as not-wanted rather than as disproven. Do not resurface on parity grounds.
Imported from beads
archie-core-1786637496982-273-2078b865· GitHub #294 · type feature · priority P2 · created 2026-08-08 · closed 2026-09-21