All IPFS operations now use IPFS_API (multiaddr format) instead of
separate IPFS_GATEWAY config. This fixes effect loading in Docker where
the gateway URL defaulted to localhost.
- Add tests for effect loading and IPFS configuration
- Simplify docker-compose.yml (remove IPFS_GATEWAY)
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Testing setup:
- Add pyproject.toml with mypy and pytest configuration
- Add requirements-dev.txt for development dependencies
- Create tests/ directory with test fixtures
- Add 17 unit tests for DAG transformation pipeline
Type annotations:
- Add app/types.py with TypedDict definitions for node configs
- Add typed helper functions: transform_node, build_input_name_mapping,
bind_inputs, prepare_dag_for_execution
- Refactor run_recipe to use the new typed helpers
Regression tests for today's bugs:
- test_effect_cid_key_not_effect_hash: Verifies CID uses 'cid' key
- test_source_cid_binding_persists: Verifies bound CIDs in final DAG
Run tests with: pytest tests/ -v
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
The EFFECT executor looks for config["cid"] or config["hash"],
but the transformation was setting config["effect_hash"].
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Fix all registry lookups to use "cid" instead of "hash" key
- app/routers/recipes.py: asset and effect resolution
- tasks/execute_sexp.py: effect config lookups
- server_legacy.py references (now deleted)
- Prefer IPFS CID over local hash in cache operations
- cache_service.py: import_from_ipfs, upload_content
- orchestrate.py: plan caching
- legacy_tasks.py: node hash tracking
Remove ~7800 lines of dead code:
- server_legacy.py: replaced by modular app/ structure
- tasks/*_cid.py: unused refactoring only imported by server_legacy
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
The recipe upload was returning the SHA3-256 hash from artdag's cache
instead of the IPFS CID, causing recipe lookups to fail.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Refactor to use IPFS CID as the primary content identifier:
- Update database schema: content_hash -> cid, output_hash -> output_cid
- Update all services, routers, and tasks to use cid terminology
- Update HTML templates to display CID instead of hash
- Update cache_manager parameter names
- Update README documentation
This completes the transition to CID-only content addressing.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Upload endpoint returns both CID and content_hash
- Cache manager handles both SHA3-256 hashes and IPFS CIDs
- get_by_cid() fetches from IPFS if not cached locally
- Execute tasks support :cid in addition to :hash
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Upload effects to IPFS, return CID instead of content hash
- Fetch effects from IPFS if not in local cache
- Keep local cache for fast worker access
- Support both :cid (new) and :hash (legacy) in recipes
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- New /effects/upload endpoint for uploading effect files
- Parses PEP 723 dependencies and @-tag metadata
- Lists, gets, and deletes effects by content hash
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Execute COMPOUND nodes with combined FFmpeg filter chain
- Handle TRANSFORM, RESIZE, SEGMENT filters in chain
- Migrate orchestrator to S-expression recipes (remove YAML)
- Update API endpoints to use recipe_sexp parameter
- Extract analysis nodes from recipe for dynamic analysis
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Show ipfs://... link next to recipe S-expression header
- Links to ipfs.io gateway for viewing the raw S-expression
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Recipe service now only handles S-expressions
- Removed yaml import and all YAML parsing code
- Plans are just node outputs - cached by content hash
- Run service looks up plans from cache, falls back to legacy dir
Code is data. Everything is S-expressions.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Plans now go through cache_manager.put() for IPFS pinning
- Returns plan_cache_id and plan_ipfs_cid in result
- Plan S-expression is content-addressed like everything else
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Orchestrator saves plan as {plan_id}.sexp (content-addressed)
- Also saves {run_id}.sexp for easy lookup by run
- Falls back to JSON for legacy plans without to_sexp_string()
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Display recipe's original S-expression when available (code is data)
- Fall back to generating S-expression from plan for legacy JSON
- Run service now prefers .sexp plan files over .json
- Add get_run_plan_sexp() for direct S-expression access
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Run card shows thumbnail previews for inputs and output
- Run detail shows output media inline (image/video/audio)
- Add audio detection (MP3, FLAC, OGG, WAV) to detect_media_type
- Add debug logging for recipe count on home page
- Add console.log debugging for DAG elements
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Handle dict inputs ({"node": "id"}) when building DAG edges
- Add normalize_inputs() to convert dict inputs to node IDs for steps
- Fix _parse_inputs to use _json.loads (correct import alias)
- Add SOURCE/EFFECT/SEQUENCE colors to node color maps
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Parse inputs JSON string to list in list_runs_by_actor
- Add status field to completed runs
- Prevents "136 files" display when inputs is a string
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Home page now shows:
- Execution runs count
- Recipes count
- Media files count
- Storage providers count
All stats require authentication.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add better debug logging for recipe filtering
- Check both uploader and owner fields when filtering by actor_id
- This handles S-expression recipes that use 'owner' field
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Use detect_media_type() with magic bytes instead of mimetypes.guess_type()
which requires file extensions. Cache files are stored by content hash
without extensions, so magic byte detection is needed.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Variable inputs can now be referenced by:
- Node ID (e.g., source_4)
- Config name (e.g., "Second Video")
- Snake case (e.g., second_video, second-video)
- Node name from def binding
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
S-expression recipes use 'owner' field while YAML uses 'uploader'.
Normalize to 'uploader' so recipe listing filter works for both formats.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Click plan nodes (in DAG or list) to see details in side panel
- URL updates to #node-{id} for direct linking
- Node detail panel shows: type, status, inputs, output, config
- Inputs can be clicked to navigate to that node
- Inputs tab now shows media previews (image/video/audio)
- Steps include config data for display
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Add "id" field to plan steps so edges connect correctly.
Previously steps only had "name" but dag_elements looked for "id",
causing edges to reference non-existent nodes.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Instead of falling through to YAML parsing (which gives confusing errors),
return a clear message that artdag.sexp is required but not installed.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Add dag_elements to the get_run endpoint render call to match
what the detail.html template expects.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Pass recipe_name through create_run to display friendly names
- Update templates to show name instead of hash
- Fall back to truncated hash if no name available
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Add format detection that correctly handles ; comments
- Import artdag.sexp parser/compiler with YAML fallback
- Add execute_step_sexp and run_plan_sexp Celery tasks
- Update recipe upload to handle both S-expr and YAML formats
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Extract username from actor_id format (@user@server)
- Set total_steps and executed from recipe nodes
- Use recipe name for display instead of hash
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Check multiple locations for nodes (nodes, dag.nodes, pipeline, steps)
and compute step_count for display in recipe list.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Build artifacts list from run output_hash and detect media type
for display in the artifacts tab.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
API clients like Python requests send Accept: */* which wasn't
matching wants_json(). Switch to checking wants_html() instead
so API clients get JSON by default.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
When a recipe run completes, save the output to the user's media
with description and source tracking.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
- Render HTML template for run detail (not just JSON)
- Get recipe name from pending_runs instead of hardcoding "dag"
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
When a DAG task completes, look up actor_id from pending_runs
(where it was saved when the run started) and include it in
run_cache. Also clean up pending_runs entry after completion.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
Inputs stored in old Redis format are JSON strings - this helper
ensures they're always returned as lists regardless of source.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
JSONB columns may return strings in some cases - explicitly parse
inputs field to ensure it's always a list.
Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>
