This resolves issue #573 by moving uwf's CAS directory from
~/.uncaged/workflow/cas/ to the shared ~/.uncaged/json-cas/ location.
Changes:
- Added getGlobalCasDir() function with UNCAGED_CAS_DIR support
- Updated createUwfStore() to use global CAS directory
- Added comprehensive test coverage (11 new tests)
- Updated all existing tests for environment isolation
- Updated documentation (CLAUDE.md, README.md)
Benefits:
- Cross-tool visibility: json-cas CLI can read uwf-created nodes
- Schema sharing: both tools access same schema registry
- Future-proofing: enables json-cas render/verbose for uwf data
Fixes#573
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Store the fully assembled prompt sent to each agent in CAS as a text
node, referenced from StepNodePayload.assembledPrompt. This enables
exact reproduction of what the agent received for debugging hallucinations.
Changes:
- workflow-protocol: StepRecord + STEP_NODE_SCHEMA add assembledPrompt field
- workflow-util-agent: AgentRunResult includes assembledPrompt, run.ts stores it
- workflow-util-agent: schemas register TEXT_SCHEMA for prompt storage
- workflow-agent-claude-code: return assembled prompt from buildClaudeCodePrompt
- workflow-agent-hermes: return assembled prompt from buildHermesPrompt
- workflow-agent-builtin: return empty prompt (no prompt assembly)
- cli-workflow: step read --prompt renders the stored prompt
- All test fixtures updated for new field
Legacy steps without assembledPrompt show 'Prompt not recorded' message.
小橘 🍊
- Enhanced Biome config with test file override for noConsole
- Applied Biome auto-fixes (8 files: formatting, template literals, optional chains)
- Updated all package repository URLs to git.shazhou.work/uncaged/workflow.git
- Added workflow-agent-claude-code to publish order in scripts/publish-all.mjs
- Added --ignore-scripts flag to publish command to bypass prepublishOnly guard
- Installed vitest in root devDependencies for test infrastructure
- Created vitest.config.ts for all 8 packages with passWithNoTests: true
- Fixed 3 test files to use vitest imports instead of bun:test
- Added test and test:ci scripts to packages missing them
- Added missing build step to .gitea/workflows/ci.yml
- Renamed CI job from 'test' to 'check' for clarity
- Created workflows/solve-issue.yaml with TDD-driven issue resolution workflow
- Registered solve-issue workflow with uwf (hash: 084YVM60BR8G6)
- Added packageManager: bun@1.3.14 to root package.json
- Added preinstall guard to block npm/pnpm/yarn
- Added prepublishOnly guard to root and all 7 public packages
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
Two fixes for 'uwf thread start solve-issue' failures:
1. json-cas 0.5.2 (npm) was missing oneOf in ALLOWED_SCHEMA_KEYS.
Published json-cas 0.5.3 with the fix, bumped all packages to ^0.5.3.
2. Semantic validator only recognized oneOf-based multi-exit schemas.
Roles using $status with enum (e.g. enum: [approved, rejected]) were
incorrectly treated as single-exit. Added isEnumMultiExit() support.
Changes:
- validate-semantic.ts: isEnumMultiExit(), getEnumStatuses(), checkSingleExitMustache()
- All package.json: @uncaged/json-cas ^0.5.2 → ^0.5.3
- validate-semantic.test.ts: 5 new enum multi-exit tests (Suite 3b)
- solve-issue-tea-worktree.test.ts: updated for current workflow structure
小橘 🍊
- Move hermes ACP integration tests to __tests__/integration/
- Add test:ci script to all packages (excludes integration/)
- CI workflow uses test:ci instead of test
- bun test still runs everything (unit + integration)
Refs #510
Refactored runBuiltinLoop function to reduce cognitive complexity from 30 to below 15 by extracting helper functions:
- shouldInjectDeadlineWarning: checks if deadline warning should be shown
- shouldProcessToolCalls: determines if tool calls should be processed
- extractFinalText: extracts last assistant message content
- injectDeadlineWarning: injects deadline warning message
- handleTextOnlyTurn: handles text-only turn logic
- handleToolCallTurn: handles tool call turn logic
- processLoopIteration: processes a single loop iteration
Added 24 new unit tests for the extracted helper functions, bringing total test count to 41 (all passing). All existing behavior is preserved.
Fixes#444
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
- Root README: add all 9 packages to table, update architecture diagram,
refresh CLI reference from uwf --help
- New READMEs for 8 packages (cli-workflow, workflow-protocol,
workflow-moderator, workflow-agent-kit, workflow-agent-hermes,
workflow-agent-builtin, workflow-agent-claude-code, workflow-dashboard)
- Updated workflow-util README to match current exports
- All API sections verified against src/index.ts exports
小橘 🍊（NEKO Team）
- Nudge turns don't consume turn budget (up to MAX_NUDGES=3), prevents
wasting agent work capacity on bookkeeping
- Inject deadline warning when 3 turns remain, telling agent to wrap up
- Agent can use status:failed to gracefully exit if it can't finish
- Inject user message when 3 turns remain, telling agent to wrap up
- Prompt tells agent to use status:failed if it can't finish in time
- Prevents wasting all turns without producing any frontmatter output
- Remove stale test file from dogfood agent run
copilot-api returns tool_calls even when tools field is omitted from
the request (infers from message history). Now the loop explicitly
nullifies tool_calls when noTools=true.
LLM sometimes emits plain text (e.g. 'Now I'll write the tests...')
without calling tools, which the loop treated as final output. Now
the loop detects this and injects a user message nudging the LLM
to either continue using tools or output frontmatter with ---.
Agent was using all continue turns to keep calling tools instead of
outputting the required frontmatter. Now continue runs with noTools=true,
forcing LLM to emit text-only response.
Also supports null tools in chatCompletionWithTools to omit tools from
the API request entirely.
Agent was wasting turns exploring the filesystem because it didn't
know its working directory. Now the system prompt includes:
'Your working directory is: /path/to/cwd'
- Add stripPreamble() to handle LLM output with text before ---
- Strengthen system prompt: CRITICAL instruction for --- at position 0
- Fixes frontmatter parsing failures on first output turn
Previously runBuiltinWithMessages deleted the session jsonl after each
run/continue call. This meant the createAgent retry mechanism (which
calls continue on frontmatter validation failure) would lose all
previous turn data — each continue started with an empty jsonl.
Now the session jsonl accumulates across run + continue calls, so the
final storeBuiltinDetail captures all turns. The jsonl file is left
behind for debugging; it's small and can be cleaned up on next startup.
Also add a workflow hint to the system prompt reminding the LLM to use
tools before outputting frontmatter, preventing premature text-only
responses on the first turn.
Each turn (assistant response / tool result) is appended to a JSONL file
at ~/.uncaged/workflow/sessions/<sessionId>.jsonl during the loop.
On completion, the JSONL is read back, each turn is stored as a CAS node,
and the detail payload references them as a flat turns[] array in
chronological order. The session file is then deleted.
Benefits:
- Real-time observability: tail -f the JSONL to watch loop progress
- Crash recovery: partial JSONL survives process death
- Zero write contention: one file per session
- Detail stays a flat array for easy consumption by CLI/dashboard
Changes:
- New session.ts: initSessionDir, appendSessionTurn, readSessionTurns, removeSession
- loop.ts: append JSONL each turn instead of accumulating in-memory
- detail.ts: reads session JSONL → persists turns to CAS → stores detail
- agent.ts: passes storageRoot/sessionId to loop, cleans up session on completion
- types.ts: remove index from TurnPayload (order is implicit in JSONL/array)
- schemas.ts: sync with type changes
Ref: #433
- StepRecord adds edgePrompt field (backward compat: defaults to "")
- StepNode CAS schema includes edgePrompt
- writeStepNode persists ctx.edgePrompt
- buildHistory exposes edgePrompt in StepContext
- buildBuiltinMessages reconstructs multi-turn moderator↔agent conversation:
system = role prompt + output format (stable prefix)
per prior visit: user (edgePrompt + inter-step summary) + assistant (output)
current: user (edgePrompt + recent summary)
- Zero extra persistence — pure function of CAS chain
- Stable prefix for LLM prompt cache hits
- 10 builtin tests pass, all other package tests pass
System message = agent identity (role prompt + output format instruction)
User message = moderator speech (task + edge prompt + history)
This reflects the workflow's core model: moderator speaks to agent
via the graph's edge prompt. Previously all content was in a single
system message with no user message, causing Claude API 400 errors.
- buildBuiltinPrompt now returns { system, user } instead of string
- agent.ts sends system + user as separate messages
- Tests updated accordingly
- Replace resolvePathInWorkspace with simple resolvePath (no boundary check)
- Remove UWF_BUILTIN_ALLOW_SHELL env gate from run_command
- Update tests accordingly
Per review: sandbox was false security with shell=true, and path
restrictions are unnecessary for a trusted agent environment.
Built-in role agent that uses workflow config models directly,
with its own tool-calling run loop. No external agent dependency.
- OpenAI-compatible chat completion client with tool_calls support
- P0 toolkit: read_file, write_file, run_command
- Integrates via createAgent factory from workflow-agent-kit
- CAS detail recording for each turn
- Path sandboxing and shell opt-in (UWF_BUILTIN_ALLOW_SHELL)
xiaomo
xiaoju
xingyue