4aa86ff1cb
[loop-cycle-150] test: add 22 unit tests for agents/base.py — BaseAgent and SubAgent ( #350 )
2026-03-18 21:10:08 -04:00
11357ffdb4
test: add comprehensive unit tests for agentic_loop.py ( #345 )
...
Co-authored-by: Kimi Agent <kimi@timmy.local >
Co-committed-by: Kimi Agent <kimi@timmy.local >
2026-03-18 20:54:02 -04:00
fcbb2b848b
test: add unit tests for jot_note and log_decision artifact tools ( #341 )
...
Co-authored-by: Kimi Agent <kimi@timmy.local >
Co-committed-by: Kimi Agent <kimi@timmy.local >
2026-03-18 20:47:38 -04:00
22e0d2d4b3
[loop-cycle-66] fix: replace language-model with inference-backend in error messages ( #334 )
2026-03-18 20:27:06 -04:00
bfd924fe74
[loop-cycle-65] feat: scaffold three-phase loop skeleton ( #324 ) ( #330 )
2026-03-18 20:11:02 -04:00
844923b16b
[loop-cycle-65] fix: validate file paths before filing thinking-engine issues ( #327 ) ( #329 )
2026-03-18 20:07:19 -04:00
9a21a4b0ff
feat: SensoryEvent model + SensoryBus dispatcher ( #318 )
...
Co-authored-by: Kimi Agent <kimi@timmy.local >
Co-committed-by: Kimi Agent <kimi@timmy.local >
2026-03-18 19:02:12 -04:00
ab71c71036
feat: time adapter — circadian awareness for Timmy ( #315 )
...
Co-authored-by: Kimi Agent <kimi@timmy.local >
Co-committed-by: Kimi Agent <kimi@timmy.local >
2026-03-18 18:47:09 -04:00
39939270b7
fix: Gitea webhook adapter — normalize events to sensory bus ( #309 )
...
Co-authored-by: Kimi Agent <kimi@timmy.local >
Co-committed-by: Kimi Agent <kimi@timmy.local >
2026-03-18 18:37:01 -04:00
234187c091
fix: add periodic memory status checks during thought tracking ( #311 )
...
Co-authored-by: Kimi Agent <kimi@timmy.local >
Co-committed-by: Kimi Agent <kimi@timmy.local >
2026-03-18 18:26:53 -04:00
f4106452d2
feat: implement v1 API endpoints for iPad app ( #312 )
...
Co-authored-by: manus <manus@timmy.local >
Co-committed-by: manus <manus@timmy.local >
2026-03-18 18:20:14 -04:00
f5a570c56d
fix: add real-time data disclaimer to welcome message ( #304 )
2026-03-18 16:56:21 -04:00
rockachopa
96e7961a0e
fix: make confidence visible to users when below 0.7 threshold ( #259 )
...
Co-authored-by: rockachopa <alexpaynex@gmail.com >
Co-committed-by: rockachopa <alexpaynex@gmail.com >
2026-03-15 19:36:52 -04:00
bcbdc7d7cb
feat: add thought_search tool for querying Timmy's thinking history ( #260 )
...
Co-authored-by: Kimi Agent <kimi@timmy.local >
Co-committed-by: Kimi Agent <kimi@timmy.local >
2026-03-15 19:35:58 -04:00
80aba0bf6d
[loop-cycle-63] feat: session_history tool — Timmy searches past conversations ( #251 ) ( #258 )
2026-03-15 15:11:43 -04:00
dd34dc064f
[loop-cycle-62] fix: MEMORY.md corruption and hot memory staleness ( #252 ) ( #256 )
2026-03-15 15:01:19 -04:00
7bc355eed6
[loop-cycle-61] fix: strip think tags and harden fact parsing ( #237 ) ( #254 )
2026-03-15 14:50:09 -04:00
f9911c002c
[loop-cycle-60] fix: retry with backoff on Ollama GPU contention ( #70 ) ( #238 )
2026-03-15 14:28:47 -04:00
7f656fcf22
[loop-cycle-59] feat: gematria computation tool ( #234 ) ( #235 )
2026-03-15 14:14:38 -04:00
8c63dabd9d
[loop-cycle-57] fix: wire confidence estimation into chat flow ( #231 ) ( #232 )
2026-03-15 13:58:35 -04:00
a50af74ea2
[loop-cycle-56] fix: resolve 5 lint errors on main ( #203 ) ( #224 )
2026-03-15 13:40:40 -04:00
b4cb3e9975
[loop-cycle-54] refactor: consolidate three memory stores into single table ( #37 ) ( #223 )
2026-03-15 13:33:24 -04:00
4a68f6cb8b
[loop-cycle-53] refactor: break circular imports between packages ( #164 ) ( #193 )
2026-03-15 12:52:18 -04:00
b3840238cb
[loop-cycle-52] feat: response audit trail with inputs, confidence, errors ( #144 ) ( #191 )
2026-03-15 12:34:48 -04:00
96c7e6deae
[loop-cycle-52] fix: remove all qwen3.5 references ( #182 ) ( #190 )
2026-03-15 12:34:21 -04:00
766add6415
[loop-cycle-52] test: comprehensive session_logger.py coverage ( #175 ) ( #187 )
2026-03-15 12:26:50 -04:00
e8dd065ad7
[loop-cycle-51] perf: mock subprocess in slow introspection test ( #172 ) ( #184 )
2026-03-15 12:17:50 -04:00
5b57bf3dd0
[loop-cycle-50] fix: agent retry uses exponential backoff instead of fixed 1s delay ( #174 ) ( #181 )
2026-03-15 12:08:30 -04:00
bcd6d7e321
[loop-cycle-50] refactor: replace bare sqlite3.connect() with context managers batch 2 ( #157 ) ( #180 )
2026-03-15 11:58:43 -04:00
ca01ce62ad
[loop-cycle-49] fix: mock _warmup_model in agent tests to prevent Ollama network calls ( #159 ) ( #177 )
2026-03-15 11:46:20 -04:00
f15ad3375a
[loop-cycle-47] feat: add confidence signaling module ( #143 ) ( #161 )
2026-03-15 11:20:30 -04:00
466db7aed2
[loop-cycle-44] refactor: remove dead code batch 2 — agent_core + test_agent_core ( #147 ) ( #150 )
2026-03-15 10:22:41 -04:00
d2c51763d0
[loop-cycle-43] refactor: remove 1035 lines of dead code ( #136 ) ( #146 )
2026-03-15 10:10:12 -04:00
16b31b30cb
fix: shell hand returncode bug, delete worthless python-exec test ( #140 )
...
- Fixed `proc.returncode or 0` bug that masked non-zero exit codes
- Deleted test_run_python_expression — Timmy does not run python, test was environment-dependent garbage
- Fixed test_run_nonzero_exit to use `ls` on nonexistent path instead of sys.executable
1515 passed, 76.7% coverage.
Co-authored-by: Kimi Agent <kimi@timmy.local >
Reviewed-on: http://localhost:3000/rockachopa/Timmy-time-dashboard/pulls/140
Co-authored-by: hermes <hermes@timmy.local >
Co-committed-by: hermes <hermes@timmy.local >
2026-03-15 09:56:50 -04:00
48c8efb2fb
[loop-cycle-40] fix: use get_system_prompt() in cloud backends ( #135 ) ( #138 )
...
## What
Cloud backends (Grok, Claude, AirLLM) were importing SYSTEM_PROMPT directly, which is always SYSTEM_PROMPT_LITE and contains unformatted {model_name} and {session_id} placeholders.
## Changes
- backends.py: Replace `from timmy.prompts import SYSTEM_PROMPT` with `from timmy.prompts import get_system_prompt`
- AirLLM: uses `get_system_prompt(tools_enabled=False, session_id="airllm")` (LITE tier, correct)
- Grok: uses `get_system_prompt(tools_enabled=True, session_id="grok")` (FULL tier)
- Claude: uses `get_system_prompt(tools_enabled=True, session_id="claude")` (FULL tier)
- 9 new tests verify formatted model names, correct tier selection, and session_id formatting
## Tests
1508 passed, 0 failed (41 new tests this cycle)
Fixes #135
Co-authored-by: Kimi Agent <kimi@timmy.local >
Reviewed-on: http://localhost:3000/rockachopa/Timmy-time-dashboard/pulls/138
Reviewed-by: rockachopa <alexpaynex@gmail.com >
Co-authored-by: hermes <hermes@timmy.local >
Co-committed-by: hermes <hermes@timmy.local >
2026-03-15 09:44:43 -04:00
d48d56ecc0
[loop-cycle-38] fix: add soul identity to system prompts ( #127 ) ( #134 )
...
Co-authored-by: hermes <hermes@timmy.local >
Co-committed-by: hermes <hermes@timmy.local >
2026-03-15 09:42:57 -04:00
76df262563
[loop-cycle-38] fix: add retry logic for Ollama 500 errors ( #131 ) ( #133 )
...
Co-authored-by: hermes <hermes@timmy.local >
Co-committed-by: hermes <hermes@timmy.local >
2026-03-15 09:38:21 -04:00
92e123c9e5
[loop-cycle-36] fix: create soul.md and wire into system context ( #125 ) ( #130 )
2026-03-15 08:37:24 -04:00
466ad08d7d
[loop-cycle-34] fix: mock Ollama model resolution in create_timmy tests ( #121 ) ( #126 )
2026-03-15 08:20:00 -04:00
cf48b7d904
[loop-cycle-1] fix: lint errors — ambiguous vars + unused import ( #123 ) ( #124 )
2026-03-15 08:07:19 -04:00
66544d52ed
feat: workspace heartbeat monitoring for thinking engine ( #28 )
...
- Add src/timmy/workspace.py: WorkspaceMonitor tracks correspondence.md
line count and inbox file list via data/workspace_state.json
- Wire workspace checks into _gather_system_snapshot() so Timmy sees
new workspace activity in his thinking context
- Add 'workspace' seed type for workspace-triggered reflections
- Add _check_workspace() post-hook to mark items as seen after processing
- 16 tests covering detection, mark_seen, persistence, edge cases
2026-03-14 21:51:36 -04:00
a57fd7ea09
[loop-cycle-30] fix: gitea-mcp binary name + test stabilization
...
1. gitea-mcp → gitea-mcp-server (brew binary name). Fixes Timmy's
Gitea triage — MCP server can now be found on PATH.
2. Mark test_returns_dict_with_expected_keys as @pytest.mark.slow —
it runs pytest recursively and always exceeds the 30s timeout.
3. Fix ruff F841 lint in test_cli.py (unused result= variable).
2026-03-14 21:32:39 -04:00
750659630b
policy: enforce PR-only merges to main + fix broken repl tests
...
Branch protection enabled on Gitea: direct push to main now rejected.
AGENTS.md updated with Merge Policy section documenting the workflow.
Also fixes bbbbdcd
2026-03-14 21:14:34 -04:00
b9b78adaa2
perf: eliminate redundant LLM calls in agentic loop ( #24 )
...
Three optimizations to the agentic loop:
1. Cache loop agent as singleton (avoid repeated warmups)
2. Sliding window for step context (last 2 results, not all)
3. Replace summary LLM call with deterministic summary
Saves 1 full LLM inference call per agentic loop invocation
(30-60s on local models) and reduces context window pressure.
Also fixes pre-existing test_cli.py repl test bugs (missing result= assignment).
2026-03-14 20:55:52 -04:00
bbbbdcdfa9
fix: remove unused variable in repl test
2026-03-14 20:45:25 -04:00
65e5e7786f
feat: REPL mode, stdin support, multi-word fix for CLI ( #26 )
2026-03-14 20:45:25 -04:00
547b502718
fix: smart_read_file accepts path= kwarg from LLMs ( #113 )
...
LLMs naturally call read_file(path=...) but the wrapper only accepted
file_name=. Pydantic strict validation rejected the mismatch. Now accepts
both file_name and path kwargs, with clear error on missing both.
Added 6 tests covering: positional args, path kwarg, no-args error,
directory listing, empty dir, hidden file filtering.
2026-03-14 20:40:19 -04:00
3e7a35b3df
Merge pull request '[loop-cycle-12] feat: Kimi delegation tool for coding tasks ( #67 )' ( #112 ) from fix/kimi-delegation-67 into main
2026-03-14 20:31:08 -04:00
453c9a0694
feat: add delegate_to_kimi() tool for coding delegation ( #67 )
...
Timmy can now delegate coding tasks to Kimi CLI (262K context).
Includes timeout handling, workdir validation, output truncation.
Sovereign division of labor — Timmy plans, Kimi codes.
2026-03-14 20:29:03 -04:00
2fb104528f
feat: add run_self_tests() tool for self-verification ( #65 )
...
Timmy can now run his own test suite via the run_self_tests() tool.
Supports 'fast' (unit only), 'full', or specific path scopes.
Returns structured results with pass/fail counts.
Sovereign self-verification — a fundamental capability.
2026-03-14 20:28:24 -04:00
