Charles YaoandClaude Opus 5
4082cceb24
The two semantic guards describe defects the ENGINE ACCEPTS: a new branch wired into a variable-aggregator that was never told about it, an array re-sent with an element quietly dropped. Both drafts start, both run green, and neither produces a `DraftWouldNotStartError` for `apply_repair` to veto. So the guards had one hole left, and it was in the agent's own fallback: `build_edit_intents` spent its corrective re-prompt on the guard's reason and then, if the retry yielded nothing usable, returned the FIRST attempt's applicable intents -- the very batch the guard had condemned. The guard fired and the silently-wrong draft was written anyway. `vet_intents` now carries `would_run_wrong` alongside `rejections`: the subset of refusals the engine does not make. The two kinds fail differently and a caller has to tell them apart without reading the text -- a structural refusal DROPPED its intent, so what survives is a batch nobody judged wrong and keeping it is a deliberate choice; a semantic verdict drops nothing and condemns what remains. Reading that off the prose would be the same mistake as keying control flow on the model's prose. `build_edit_intents` therefore raises `ProposalWouldRunWrongError` instead of returning a batch still carrying one, whichever attempt it came from. Nothing is written and nothing is attempted. `handle_plan_approval` catches it into the same `_change_not_applied` recovery a refused write gets: the plan stays at its gate, the reason goes on an error card and into `last_edit_rejection` for the next attempt. Surfacing is a real exit now and not the dead end it would have been before this plan: Task 7 gave the gate a revert and a keep-adjusting action, and Task 6 carries the refusal into the next proposal, so the user can change the rules and the retry is not blind. `ProposalWouldRunWrongError` is deliberately neither a `DraftWouldNotStartError` nor a `ValueError`: the draft WOULD start, and a broad `except ValueError` around a port write must not swallow it as if the write had been attempted. `_change_not_applied` takes the activity to mark failed, defaulting to the write where the usual refusal happens. This exit gives up during "prepare", so failing the write would have shown a LATER step failed while an earlier one was still running, with the step between never resolving. Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Dify Backend API
Setup and Run
Important
In the v1.3.0 release,
poetryhas been replaced withuvas the package manager for Dify API backend service.
uv and pnpm are required to run the setup and development commands below.
Using scripts (recommended)
The scripts resolve paths relative to their location, so you can run them from anywhere.
-
Run setup (copies env files and installs dependencies).
./dev/setup -
Review
api/.env,web/.env.local, anddocker/middleware.envvalues (see theSECRET_KEYnote below). -
Start middleware (PostgreSQL/Redis/Weaviate).
./dev/start-docker-compose -
Start backend (runs migrations first).
./dev/start-api -
Start Dify web service.
./dev/start-web./dev/setupand./dev/start-webinstall JavaScript dependencies through the repository root workspace, so you do not need a separatecd web && pnpm installstep. -
Set up your application by visiting
http://localhost:3000. -
Start the worker service (async and scheduler tasks, runs from
api)../dev/start-worker -
Optional: start Celery Beat (scheduled tasks).
./dev/start-beat
Environment notes
Important
When the frontend and backend run on different subdomains, set COOKIE_DOMAIN to the site’s top-level domain (e.g.,
example.com). The frontend and backend must be under the same top-level domain in order to share authentication cookies.
-
Generate a
SECRET_KEYin the.envfile.bash for Linux
sed -i "/^SECRET_KEY=/c\\SECRET_KEY=$(openssl rand -base64 42)" .envbash for Mac
secret_key=$(openssl rand -base64 42) sed -i '' "/^SECRET_KEY=/c\\ SECRET_KEY=${secret_key}" .env
Testing
-
Install dependencies for both the backend and the test environment
cd api uv sync --group dev -
Run the tests locally with mocked system environment variables in
tool.pytest_envsection inpyproject.toml, more can check AGENTS.mdContinue in the
apidirectory from the previous step.uv run pytest # Run all tests uv run pytest tests/unit_tests/ # Unit tests only uv run pytest tests/integration_tests/ # Integration tests # Code quality ../dev/reformat # Run all formatters and linters uv run ruff check --fix ./ # Fix linting issues uv run ruff format ./ # Format code uv run pyrefly check # Type checking
Generate TS stub
uv run dev/generate_swagger_specs.py --output-dir openapi
use https://jsontotable.org/openapi-to-typescript to convert to typescript