Commit Graph
8 Commits
Author SHA1 Message Date
jiang addc81e106 fix(agent): align recovery reports with completed incident state
Generic Container CI/CD / test-build-publish (push) Successful in 57s
Agent CI/CD / build-test-publish-and-deploy (push) Successful in 57s
The investigation instructions previously separated analysis and field work without explaining how passed recovery evidence completes an incident. Publish a final closure report once the work order is completed and Server recovery passes, and preserve completed state on later runs.

Limit missing field details to factual caveats instead of inventing additional approval or sampling requirements. Update the investigation prompt version and descriptor.
2026-11-08 04:31:11 +08:00
jiang 874bfba907 fix: harden report tools and guide investigation presentation
Generic Container CI/CD / test-build-publish (push) Successful in 56s
Agent CI/CD / build-test-publish-and-deploy (push) Successful in 56s
2026-11-02 15:23:34 +08:00
jiang 6336157ff4 feat: upgrade investigation and presentation runtime
Generic Container CI/CD / test-build-publish (push) Successful in 1m6s
Agent CI/CD / build-test-publish-and-deploy (push) Successful in 1m7s
2026-10-30 17:30:19 +08:00
jiang 52681a8451 fix: preserve interaction replies across chat recovery 2026-10-29 17:53:33 +08:00
jiang 7d47a2df86 feat: add scoped history and artifacts
Generic Container CI/CD / test-build-publish (push) Failing after 8m25s
Agent CI/CD / build-test-publish-and-deploy (push) Failing after 8m25s
2026-10-22 18:40:20 +08:00
jiang 6a49d2209e feat(agent): unify MCP tools and adopt Responses delivery
Move project and host queries into native MCP with host-owned OAuth/context,
while retaining memory, skills, actions and publication plugins. Add nearby-model
queries and explicit chat/task/learning model defaults.

Adopt Responses HTTP/SSE and native steer/queue inbox delivery, preserving scoped
session ownership, pending input edits and investigation communication.

Disable local-provider discovery: Server /health allowed native vLLM to probe
/v1/models on port 8000. A native before/after regression prevents recurrence.

Validation: 170 focused Bun tests; service/plugin typechecks; native discovery,
provider-discovery and input-inbox smoke checks; no paid model inference.
2026-10-19 18:46:58 +08:00
jiang f5c3cde2c2 feat(chat): stream and persist local answer UI with final report references 2026-10-12 18:54:18 +08:00
jiang d58bd19039 feat(agent): unify models and validate OpenCode providers
Generate chat, task and learning models from agent-models.json.
Use WaterNetwork fast/expert models with 272k context and validate
upstream names before readiness. Add an isolated model check command
with signal-safe cleanup and plugin, Skill and MCP management docs.

Verified in isolation: type checks, 26 focused tests and both model
catalog checks; no model inference requests.
2026-10-09 11:28:37 +08:00