Renamed skill files from using colons to dashes (e.g., tessl:add-api-endpoint → tessl-add-api-endpoint) to fix checkout issues on Windows filesystems which don't allow colons in filenames.
Skills will continue to work as the files contain relative paths to .tessl directory and don't reference their own filenames.
The LLM prompt preprocessing now highlights __init__ signatures for
regular classes, not just @dataclass ones, reducing brute-force
constructor guessing and pytest.skip() fallbacks in generated tests.
log_features() appended test results in call-completion order, causing
model attribution swaps when LLM responses arrived out of order. Pass
test_index through and use positional insertion instead of append.
Replace sticky positioning + ResizeObserver height calc with a flex
column layout (h-screen container, flex-1 panel group) that reliably
fills the viewport. Drop useDefaultLayout hook (not SSR-safe) in favor
of manual localStorage persistence inside useEffect.
Replace the fixed 480px chat overlay with a draggable split-pane layout
using react-resizable-panels, and make tool rounds expandable to show
the actual data the agent retrieved (code, errors, LLM call details).
Remove MAX_TOOL_ROUNDS cap so the model decides when to stop calling
tools. Add a safety net that makes a final tool-free API call if the
loop ends without emitting any visible text, fixing empty assistant
bubbles. Clean up redundant comments.
- Move trace data to top of prompt (long-context best practice: data
before instructions improves quality ~30%)
- Wrap sections in XML tags (<trace_data>, <role>, <domain_knowledge>,
<guidelines>, <use_parallel_tool_calls>) for better parseability
- Remove aggressive language (MUST, CRITICAL, HARD REQUIREMENT) that
causes overtriggering on Opus 4.6
- Replace rigid 4-step investigation workflow with general guidelines
to let adaptive thinking handle reasoning strategy
- Remove duplicate content (tool reference section, two checklists)
- Add <use_parallel_tool_calls> block per Anthropic's recommended pattern
- Tone down tool descriptions from directive to descriptive
- Net reduction: 49 fewer lines in system prompt
Thinking blocks from previous tool rounds (10-50KB each) were
accumulating in conversation history, causing Azure AI Foundry to hang
after 4+ rounds. Redact thinking content before each API call while
preserving required block structure. Also adds per-round timeout safety
net and status indicators between rounds.
Eliminate redundant API call by extracting text from the loop's final
response directly instead of making a separate streaming call. Pre-build
candidatesBySource, candidatesById, and testModelMap in indexTraceData()
to replace repeated O(n) linear searches in tool calls and prompt
building. Combine cost/token aggregation into a single pass.
Restructure agent loop to use stream()+finalMessage() for all API calls,
fixing the SDK's non-streaming timeout error with max_tokens 32k. Add
parallel tool execution, tool activity bubbles in the frontend, and
restructure the system prompt for better investigation behavior.
Guide the chat agent to use the new tools proactively: a DEBUGGING TOOLS
section with structured guidance for get_llm_call_detail and codebase
browsing, a 4-step workflow (OBSERVE → INVESTIGATE → LOCATE → RECOMMEND),
and a RESPONSE CHECKLIST at the end of the prompt requiring the agent to
cite real file paths before responding.
Give the observability chat agent four new tools: get_llm_call_detail
(full prompt/response for any LLM call), read_file, search_code, and
list_directory for navigating the codeflash-internal and codeflash CLI
repos. This lets the agent trace problems end-to-end from trace data
through actual prompts to pipeline source code.
- Add id to IndexedTraceData.llmCalls so the agent can reference calls
- Make resolveToolCall async (Prisma + fs + child_process)
- Make processToolUseResponse async to match
- Bump MAX_TOOL_ROUNDS from 5 to 15 for multi-step code browsing
- Add CODEFLASH_INTERNAL_REPO_PATH / CODEFLASH_CLI_REPO_PATH env vars
- Path traversal protection, file size caps, search result limits
## Summary
- Add Foundry env vars (ANTHROPIC_FOUNDRY_API_KEY,
ANTHROPIC_FOUNDRY_BASE_URL) so the workflow authenticates via Azure
Foundry
- Fix Serena language config (javascript -> typescript)
## Summary
- Adds the GitHub Agentic Workflows duplicate code detector, configured
for Python and TypeScript/JavaScript with Serena semantic analysis
- Runs daily, flags patterns spanning 10+ lines or appearing in 3+
locations
- Creates up to 3 issues per run with `[duplicate-code]` prefix
## Notes
- Requires Claude API secret configured in repo Actions secrets
- `code-quality` and `automated-analysis` labels will be auto-created on
first run
## Summary
- Restructure CLAUDE.md hierarchy so Claude Code auto-discovers
project-specific instructions
- Delete dead `AGENTS.md` files (referenced non-existent
`.tessl/RULES.md`)
- Rename `django/aiservice/AGENTS.md` → `CLAUDE.md` for auto-discovery
- Create `js/CLAUDE.md` with package commands and gotchas
- Move PR review guidelines to `.claude/rules/pr-review.md` (auto-loaded
rule)
- Move prek workflow to `.claude/skills/fix-prek.md` (on-demand skill)
- Add path-scoped rules for Python and Next.js patterns
- Add domain glossary, service architecture diagram, and per-package
gotchas
## Test plan
- Verify `CLAUDE.md` files exist at root, `django/aiservice/`, and `js/`
- Verify no remaining references to `AGENTS.md` or `.tessl/`
- Verify `.claude/rules/` and `.claude/skills/` files are committed
## Summary
- Extract testgen and optimizer API routers from
`core/languages/python/` into `core/shared/` with lazy imports,
eliminating cross-module coupling between language modules
- Delete stale JavaScript prompt files left in the Python module after
migration to `js_ts/`
- Remove backward-compat fallback paths for prompt files that already
exist at their new locations
- Remove unused `is_multi_context_any()` and its cross-language imports
- Remove unused `BEGIN_PATCH`/`END_PATCH` constants and stale TODO
## Test plan
- [ ] Verify testgen endpoint dispatches correctly for Python, JS/TS,
and Java
- [ ] Verify optimizer endpoint dispatches correctly for all languages
- [ ] Run existing testgen and optimizer tests
## Summary
- Pass test_index through LLM call context so observability chat can
attribute responses to specific test generation calls
- Fix SSE streaming to send keepalive pings from the start
CF-504
# Pull Request Checklist
## Description
- [ ] **Breaking Changes**: Document any breaking changes (if
applicable)
- [ ] **Description of PR**: Clear and concise description of what this
PR accomplishes
- [ ] **Related Issues**: Link to any related issues or tickets
## Testing
- [ ] **Test cases Attached**: All relevant test cases have been
added/updated
- [ ] **Manual Testing**: Manual testing completed for the changes
## Monitoring & Debugging
- [ ] **Logging in place**: Appropriate logging has been added for
debugging user issues
- [ ] **Sentry will be able to catch errors**: Error handling ensures
Sentry can capture and report errors
- [ ] **Avoid Dev based/Prisma logging**: No development-only or
Prisma-specific logging in production code
## Configuration
- [ ] **Env variables newly added**: Any new environment variables are
documented in .env.example file or mentioned in description
---
## Additional Notes
<!-- Add any additional context, screenshots, or notes for reviewers
here -->
---------
Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>
Co-authored-by: claude[bot] <209825114+claude[bot]@users.noreply.github.com>
Co-authored-by: HeshamHM28 <HeshamMohamedFathy@outlook.com>
Co-authored-by: Ubuntu <ubuntu@ip-172-31-39-200.ec2.internal>
Co-authored-by: claude[bot] <41898282+claude[bot]@users.noreply.github.com>
Co-authored-by: Kevin Turcios <turcioskevinr@gmail.com>
Co-authored-by: Kevin Turcios <106575910+KRRT7@users.noreply.github.com>
## Problem
The JS/TS language handler (`core/languages/js_ts/`) was importing
models, schemas, config, prompts, and helpers directly from the Python
language handler. This created a confusing architectural dependency and
risked serving wrong language-specific prompt content.
## What Changed
- Created `core/shared/` for genuinely language-agnostic code (optimizer
schemas, models, config, testgen models, context helpers)
- Moved JS/TS-specific prompts and context helpers into
`core/languages/js_ts/`
- Updated all consumers (20+ files) to import from the correct locations
- Removed backwards-compat re-exports from the Python module
## Result
- **Before:** 11 imports from `core.languages.python` in
`core/languages/js_ts/`
- **After:** 0
## Summary
- Chat panel on the observability timeline that uses Claude to answer
questions about optimization traces
- Tool-based context retrieval (fetches candidates, tests, errors on
demand instead of stuffing everything upfront)
- Uses `@anthropic-ai/sdk` via Azure AI Foundry
- Strengthened testgen prompts to ban mocks/fakes for test inputs
## Summary
- Use claude-opus-4-6 model for both pr-review and claude-mention jobs
- Add mypy checks and consolidated summary comment (Steps 1 & 4) from
CLI workflow
- Add Edit tool and extra git/gh tools to allowed tools
