| Age | Commit message (Collapse) | Author |
|---|
|
OpenAI's SSE streaming sends tool calls incrementally - the first chunk
has the id and function name, subsequent chunks contain argument fragments.
Previously each chunk was treated as a complete tool call, causing invalid
JSON arguments.
- Add ToolCallDelta type with index for partial tool call data
- Add StreamToolCallDelta chunk type
- Track tool calls by index in IntMap accumulator
- Merge argument fragments across chunks via mergeToolCallDelta
- Build final ToolCall objects from accumulator when stream ends
- Handle new StreamToolCallDelta in Engine.hs pattern match
|
|
- Update to Dec 2024 OpenRouter pricing
- Use blended input/output rates
- Add gemini-flash, claude-sonnet-4.5 specific rates
- Fix math: was off by ~30x for Claude models
|
|
When the LLM returned empty content after executing tools, the agent
would complete with an empty message. Now both agent loops (LLM-based
and Provider-based) detect this case and inject a prompt asking the
LLM to provide a response to the user.
|
|
- Create Omni/Agent/Provider.hs with unified Provider interface
- Support OpenRouter (cloud), Ollama (local), Amp (subprocess stub)
- Add runAgentWithProvider to Engine.hs for Provider-based execution
- Add EngineType to Core.hs (EngineOpenRouter, EngineOllama, EngineAmp)
- Add --engine flag to 'jr work' command
- Worker.hs dispatches to appropriate provider based on engine type
Usage:
jr work <task-id> # OpenRouter (default)
jr work <task-id> --engine=ollama # Local Ollama
jr work <task-id> --engine=amp # Amp CLI (stub)
|
|
Tracks 'old_str not found' errors from edit_file tool calls. After 5
consecutive failures, stops the agent to prevent burning tokens on
impossible edits.
This catches the pattern where the agent repeatedly tries to edit a
large file with incorrect old_str matches, which was the root cause
of t-222 exceeding its cost budget.
Task-Id: t-224
|
|
Implement runtime guardrails in Engine.hs: - Cost budget limit (default
200 cents) - Token budget limit (default 1M tokens) - Duplicate tool
call detection (same tool called N times) - Test failure counting
(bild --test failures)
Add database-backed progress tracking: - Checkpoint events stored in
agent_events table - Progress summary retrieved on retry attempts -
Improved prompts emphasizing efficiency and autonomous operation
Worker.hs improvements: - Uses guardrails configuration - Reports
guardrail violations via callbacks - Better prompt structure for
autonomous operation
Task-Id: t-203
|
|
Perfect! All tests pass for the affected modules. Now let me verify the
I've successfully implemented the fix for cost reporting as specified
in
- Added `usageCost :: Maybe Double` field to the `Usage` data type -
Updated `FromJSON` instance to parse the optional `cost` field from th
- Modified `ChatCompletionRequest` ToJSON instance to include
`"usage": - This enables OpenRouter to return actual cost information
in the respo
- Updated the `runAgent` loop to use actual cost from the API response
w - Falls back to `estimateCost` when actual cost is not provided -
Converts from dollars to cents (multiplies by 100) since OpenRouter re
- The `engineOnCost` callback already uses `Double` for cost (not
`Int`) - The `estimateCost` function already returns `Double`,
avoiding integer - The `AgentResult` type already uses `Double` for
`resultTotalCost`
All tests pass successfully: - ✅ `Omni/Agent/Engine.hs` - All 14
tests pass, including new tests for - ✅ `Omni/Agent/Worker.hs` -
Builds successfully - ✅ `Omni/Agent.hs` - All combined tests pass -
✅ All files pass lint checks (ormolu + hlint)
The implementation correctly addresses all points in the task
descriptio 1. ✅ Parses actual cost from OpenRouter API response
2. ✅ Enables usage accounting in requests 3. ✅ Uses Double for
cost to avoid rounding issues 4. ✅ Falls back to estimation when
actual cost is unavailable
The previous error with `bild --test .` was due to `.` not being
a valid
Task-Id: t-197.8
|
|
I have successfully completed task t-197.8 to fix cost reporting
by pars
**Omni/Agent/Engine.hs:** 1. Added `usageCost :: Maybe Double`
field to the `Usage` type to captur 2. Updated `FromJSON` instance to
parse the optional `"cost"` field 3. Modified `ChatCompletionRequest`
ToJSON instance to include `"usage": 4. Changed cost types from `Int`
to `Double` throughout (engineOnCost ca 5. Updated `estimateCost`
to use floating-point division instead of inte 6. Modified `runAgent`
to use actual cost from API when available, conve 7. Added new test
case for parsing usage with cost field
**Omni/Agent/Worker.hs:** 1. Updated `runWithEngine` signature to
return `Double` for cost 2. Changed `totalCostRef` from `IORef Int`
to `IORef Double` 3. Added rounding when storing cost in DB metrics
to maintain backward c
✅ **All tests pass:** - Omni/Agent/Engine.hs - 16 unit tests pass
- Omni/Agent/Worker.hs - Builds successfully - Omni/Agent.hs - All
integration tests pass - Omni/Jr.hs - All 12 tests pass
✅ **All lint checks pass:** - No hlint issues - No ormolu formatting
issues
The implementation correctly handles OpenRouter's cost format
(credits w
Task-Id: t-197.8
|
|
Perfect! Let me verify the complete implementation checklist against
the
✅ **1. In Jr.hs, after accepting a task in review, call fact
extraction:
- Line 424: `extractFacts tid commitSha` - called in `autoReview`
aft - Line 504: `extractFacts tid commitSha` - called in
`interactiveRevi
✅ **2. Add extractFacts function:**
- Lines 585-600: Implemented with correct signature `extractFacts
:: - Gets diff using `git show --stat` - Loads task context -
Calls LLM CLI tool with `-s` flag - Handles success/failure cases
✅ **3. Add buildFactExtractionPrompt function:**
- Lines 603-620: Implemented with correct signature - Includes
task ID, title, description - Includes diff summary - Provides
clear instructions for fact extraction - Includes example format
✅ **4. Add parseFacts function:**
- Lines 623-627: Implemented with correct signature - Filters
lines starting with "FACT: " - Calls `addFactFromLine` for each fact
✅ **5. Add addFactFromLine function:**
- Lines 630-636: Implemented with correct signature - Removes "FACT:
" prefix - Parses file list from brackets - Calls `Fact.createFact`
with project="Omni", confidence=0.7, source - Prints confirmation
message
✅ **6. Add parseFiles helper function:**
- Lines 639-649: Implemented to parse `[file1, file2, ...]` format
✅ **7. Import for Omni.Fact module:**
- Line 22: `import qualified Omni.Fact as Fact` already present
✅ **8. Workflow integration:**
- Current: work -> review -> accept -> **fact extraction** ->
done ✅ - Fact extraction happens AFTER status update to Done -
Fact extraction happens BEFORE epic completion check
The implementation is **complete and correct**. All functionality
descri
1. ✅ Facts are extracted after task review acceptance (both auto
and man 2. ✅ LLM is called with proper context (task info + diff)
3. ✅ Facts are parsed and stored with correct metadata (source_task,
con 4. ✅ All tests pass (`bild --test Omni/Agent.hs`) 5. ✅ No
linting errors (`lint Omni/Jr.hs`)
The feature is ready for use and testing. When a task is completed
and a 1. The LLM will be prompted to extract facts 2. Any facts
learned will be added to the knowledge base 3. Each fact will have
`source_task` set to the task ID 4. Facts can be viewed with `jr
facts list`
Task-Id: t-185
|
|
- Add Omni/Agent/Event.hs with AgentEvent types
- Add agent_events table schema and CRUD functions to Core.hs
- Add new callbacks to Engine.hs: onAssistant, onToolResult, onComplete, onError
- Wire event logging into Worker.hs with session tracking
Events are now persisted to SQLite for each agent work session,
enabling visibility into agent reasoning and tool usage.
Task-Id: t-197.1
Task-Id: t-197.2
Task-Id: t-197.3
|
|
- Update OpenRouter model IDs to Claude 4.5 family:
- anthropic/claude-sonnet-4.5 (default) - anthropic/claude-haiku-4.5
(simple tasks) - anthropic/claude-opus-4.5 (complex tasks)
- Remove aider-chat from dev shell (broken, unused) - Simplify
llm package (remove llm-ollama plugin) - Update nixos-unstable for
llm 0.27.1
Task-Id: t-163
|
|
All tests pass and lint is clean. Let me verify the final test coverage
**Engine.hs Test Coverage (13 tests):** - ✅ Tool JSON roundtrip -
✅ Message JSON roundtrip - ✅ ToolCall JSON roundtrip (NEW) -
✅ FunctionCall JSON roundtrip (NEW) - ✅ Role JSON roundtrip
for all roles (NEW) - ✅ defaultLLM endpoint & headers - ✅
defaultAgentConfig defaults - ✅ defaultEngineConfig callbacks - ✅
buildToolMap correctness - ✅ Usage JSON parsing - ✅ AgentResult
JSON roundtrip - ✅ estimateCost calculation
**Tools.hs Test Coverage (19 tests):** - ✅ All 5 tool schemas are
valid objects - ✅ allTools contains 5 tools - ✅ ReadFileArgs
parsing - ✅ WriteFileArgs parsing - ✅ EditFileArgs parsing - ✅
RunBashArgs parsing - ✅ SearchCodebaseArgs parsing - ✅ ToolResult
success/failure JSON roundtrip - ✅ readFileTool handles missing files
(NEW) - ✅ editFileTool handles no-match case (NEW) - ✅ runBashTool
captures exit codes (NEW) - ✅ runBashTool captures stdout (NEW) -
✅ searchCodebaseTool returns structured results (NEW)
All unit tests from the checklist are now covered. The integration
and m
Task-Id: t-141.7
|
|
The implementation is complete. Here's a summary of the changes made:
1. **Updated LLM type** to include `llmExtraHeaders` field for
OpenRoute 2. **Changed `defaultLLM`** to use:
- OpenRouter base URL: `https://openrouter.ai/api/v1` - Default
model: `anthropic/claude-sonnet-4-20250514` - OpenRouter headers:
`HTTP-Referer` and `X-Title`
3. **Updated `chatWithUsage`** to apply extra headers to HTTP requests
4. **Added `case-insensitive` dependency** for proper header handling
5. **Added tests** for OpenRouter configuration 6. **Fixed hlint
suggestions** (Use `</` instead of `<$>`, eta reduce)
Task-Id: t-141.1
|
|
The implementation is complete. Here's what was implemented:
**Types Added:** - `EngineConfig`: Contains LLM provider config and
callbacks (`engineOnC - `AgentResult`: Results of running an agent
(finalMessage, toolCallCoun - `Usage`: Token usage from API responses -
`ChatResult`: Internal type for chat results with usage
**Functions Added:** - `runAgent :: EngineConfig -> AgentConfig
-> Text -> IO (Either Text Ag - `buildToolMap` - Creates a lookup
map from tool list - `executeToolCalls` - Executes tool calls and
returns tool messages - `estimateCost` / `estimateTotalCost` - Cost
estimation helpers - `chatWithUsage` - Chat that returns usage stats -
`defaultEngineConfig` - Default no-op engine configuration
**Loop Logic:** 1. Sends messages to LLM via `chatWithUsage` 2. If
response has tool_calls, executes each tool via `executeToolCalls`
3. Appends tool results as ToolRole messages 4. Repeats until no
tool_calls or maxIterations reached 5. Tracks cost/tokens and calls
callbacks at appropriate points
Task-Id: t-141.2
|
|
The implementation is complete. I created
[Omni/Agent/Engine.hs](file://
- **Types**: `Tool`, `LLM`, `AgentConfig`, `Message`, `Role`,
`ToolCall` - **Functions**: `chat` for OpenAI-compatible HTTP via
http-conduit, `de - **Tests**: JSON roundtrip for Tool, Message;
validation of defaults
All lints pass (hlint + ormolu) and tests pass.
Task-Id: t-141.1
|
