Compare commits

..

2 Commits

Author SHA1 Message Date
6bded2eece docs: record DAVE persona release
All checks were successful
Codex Template Compliance / template-compliance (pull_request) Successful in 9s
Build / test-and-image (pull_request) Successful in 26s
2026-07-05 22:32:37 +02:00
c07a65231f Merge pull request 'feat: add DAVE adaptive synthetic persona' (#68) from codex/issue-67-dave-persona into codex/production-intelligence-terminal
All checks were successful
Codex Template Compliance / template-compliance (push) Successful in 10s
Release Dry Run / release-dry-run (push) Successful in 19s
Build / test-and-image (push) Successful in 35s
2026-07-05 20:30:46 +00:00

View File

@@ -33,6 +33,9 @@ Last updated: 2026-07-05
- Issue #64 fixed a local-model protocol leak where an exhausted tool loop could expose a raw `tool_call` JSON object as the Telegram answer. PR #65 merged as `e47c23e685d833d5f7433dc27b0834854c8c6152`.
- Exhausted loops now switch to a separate tool-free finalization prompt, allow one bounded repair attempt, and fail closed with a localized user-facing message if the model still requests tools. Internal protocol JSON is never returned as the answer.
- PR runs 895-896 and production runs 897-899 passed. The registry `latest` image was deployed to Dockge; live health reported `running/healthy`, 14 tools enabled, no sweep error, and the encrypted Security Manager profile preserved (`profileExists=true`).
- Issue #67 / PR #68 added DAVE as the consistent synthetic Security Manager identity across the tool agent, tool-free finalizer, provider-only chat, and first-start language prompt. Production merge: `c07a65231ffc6b4d2f0c823b91aaf2eb13900e05`.
- DAVE adapts language, formality, directness, verbosity, sentence length, formatting, and technical depth from the newest message plus bounded chat history. It does not imitate errors, hostility, panic, discriminatory language, or unsupported certainty, and it never claims human consciousness, emotions, a body, or access beyond allowlisted tools.
- PR runs 905-906 and production runs 907-909 passed. The new registry image was deployed to Dockge; live health remained `running/healthy`, the 14-tool agent was enabled, no sweep error was present, and the encrypted profile persisted.
## Repository State
@@ -473,6 +476,8 @@ d13652a merge: controlled Telegram terminal agent (PR #60)
0c7ddc5 feat: add privacy-aware security manager onboarding
b42b393 fix: prevent agent tool protocol leakage
e47c23e merge: prevent Telegram agent protocol leakage (PR #65)
994c806 feat: add DAVE adaptive synthetic persona
c07a652 merge: add DAVE adaptive synthetic persona (PR #68)
```
The large implementation commit `85f97bb` and the dashboard/action fix `53470cc` are contained in both: