api_routes
diapason.server.api_routes
¶
Extended API routes for agents, workflows, memory, traces, etc.
create_agent(req, request)
async
¶
Spawn a new agent.
feedback_stats(request)
async
¶
Get feedback statistics.
get_budget(request)
async
¶
Get current budget usage and limits.
get_optimize_run(run_id, request)
async
¶
Get optimization run details.
get_session(session_id, request)
async
¶
Get a specific session.
get_trace(trace_id, request)
async
¶
Get a specific trace by ID.
include_all_routes(app)
¶
Include all extended API routers in a FastAPI app.
install_skill(request)
async
¶
Install a skill (placeholder).
kill_agent(agent_id, request)
async
¶
Kill a running agent.
learning_policy(request)
async
¶
Return current routing policy configuration.
learning_stats(request)
async
¶
Return learning system statistics across all sub-policies.
list_agents(request)
async
¶
List available agent types and running agents.
list_optimize_runs(request)
async
¶
List optimization runs.
list_sessions(request, limit=20)
async
¶
List active sessions.
list_skills(request)
async
¶
List installed skills.
list_traces(request, limit=20)
async
¶
List recent traces.
memory_config(request)
async
¶
Return current memory configuration.
Reports memory as unavailable (rather than falsely claiming
backend_type: sqlite) when the native diapason_rust extension is
missing, so the UI can show the real cause instead of a healthy-looking
config that backs a silent no-op (#502).
memory_index(req, request)
async
¶
Index files from a path into memory.
memory_search(req, request)
async
¶
Search memory for relevant content.
memory_stats(request)
async
¶
Get memory backend statistics.
memory_store(req, request)
async
¶
Store content in memory.
message_agent(agent_id, req, request)
async
¶
Send a message to a running agent.
prometheus_metrics(request)
async
¶
Prometheus-compatible metrics endpoint.
remove_skill(skill_name, request)
async
¶
Remove a skill (placeholder).
set_budget_limits(req, request)
async
¶
Update budget limits.
speech_health(request)
async
¶
Check if a speech backend is available.
start_optimize_run(req, request)
async
¶
Start a new optimization run.
submit_feedback(req, request)
async
¶
Submit feedback for a trace.
telemetry_energy(request)
async
¶
Get energy monitoring data.
telemetry_stats(request)
async
¶
Get aggregated telemetry statistics.
transcribe_speech(request)
async
¶
Transcribe uploaded audio to text.
websocket_chat_stream(websocket)
async
¶
Stream chat responses over a WebSocket connection.
Accepts JSON messages of the form::
{"message": "...", "model": "...", "agent": "..."}
Sends back JSON chunks::
{"type": "chunk", "content": "..."} -- per-token streaming
{"type": "done", "content": "..."} -- final assembled response
{"type": "error", "detail": "..."} -- on failure