nledford/engineering-review-board / skills/digital-asset-management
digital-asset-management
VerifiedDigital asset management (DAM) engineering guidance for photo and video ingest, asset identity, immutable originals, derivatives, metadata, non-destructive edits, organization, smart collections, hierarchies, viewing, sharing, and validation. Use when building, changing, reviewing, or testing a photo/video catalog, media library, asset browser, editing workflow, or related API and data model. Do not use for standalone codec implementation, generic SQL/API/CSS work, or non-media document management without a DAM lifecycle.
| File | Kind | Size |
|---|---|---|
| SKILL.md | skill md | 11.2 KB |
| references/catalog-and-persistence.md | markdown | 4.2 KB |
| references/editing-and-rendering.md | markdown | 3.1 KB |
| references/metadata-and-interoperability.md | markdown | 3.3 KB |
| references/viewer-organization-and-testing.md | markdown | 3.2 KB |
references/editing-and-rendering.md
View on GitHub# Editing And Rendering
Use this reference when an application changes edits, previews, exports,
thumbnails, video proxies, rendering jobs, or cache behavior.
## Non-Destructive Contract
- Retain immutable original bytes and import provenance.
- Persist an edit recipe as a validated ordered sequence of operations. Specify
operation identifier, schema version, parameters, units, ranges, defaults, and
error behavior.
- Distinguish recipe revision, undo/redo history, asset version, generated
rendition, and export. Each has different restore and retention semantics.
- Define whether restore moves a current-version pointer or creates a new
immutable revision. Make concurrent edit conflicts explicit.
## Operation Semantics
For crop, rotate, flip, exposure, contrast, saturation, temperature/tint, auto
color, auto white balance, sharpening, or noise reduction:
- Specify operation order and whether operations commute.
- Use one documented geometry space. State orientation normalization, normalized
versus pixel coordinates, bounds, rounding, and replacement behavior.
- Preserve resolved automatic-operation parameters plus algorithm/model version.
Storing only `auto: true` makes history non-reproducible.
- State color-space, bit-depth, alpha, HDR, and profile behavior for previews and
exports. Test tolerances deliberately when pixel equality is impossible.
- For video trim/edit operations, state timeline/timebase units, keyframe limits,
stream-copy/re-encode choice, and audio/subtitle handling.
## Rendition Lifecycle
A rendition key normally reflects every input that changes output:
```text
source identity + edit recipe revision + orientation/presentation metadata
+ color/output profile + renderer/codec version + rendition specification
```
- Write generated bytes and completion state atomically enough that readers do
not receive partial output.
- Invalidate or re-authorize renditions after source, recipe, profile, renderer,
privacy, or ACL changes. Cache hits must not bypass current authorization.
- Define queue ownership, deduplication, cancellation, retry/backoff, terminal
failure, and garbage collection. Avoid unbounded concurrent decodes/renders.
- Enforce per-item and per-tenant/workload CPU, wall-time, memory, temporary-file,
and output limits. Run risky decoders/transcoders with least privilege in an
operating-system, container, or VM isolation boundary appropriate to the
threat model, without ambient credentials or unnecessary network/filesystem
access. Pass paths and parameters through structured argument APIs, not a shell.
- Define preview/full-resolution parity and allowed differences.
## Tests
- Use fixed fixtures for orientation, transparency, ICC/HDR, malformed media,
mislabeled/polyglot files, decompression/resource bombs, sidecars,
variable-frame-rate/multi-stream video, and corrupted derivatives.
- Test originals remain unchanged, restore remains traceable, automatic operations
retain resolved parameters, and changed recipes invalidate output.
- Measure bounded decode/render memory, cancellation, cache-hit/stale behavior,
and concurrent generation at realistic library sizes.
Related skills
| Skill | Stars | Updated | Works with | Status | Install |
|---|---|---|---|---|---|
| finishing-a-development-branchobra/superpowersUse when implementation is complete, all tests pass, and you need to decide how to integrate the work | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| systematic-debuggingobra/superpowersUse when encountering any bug, test failure, or unexpected behavior, before proposing fixes | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| test-driven-developmentobra/superpowersUse when implementing any feature or bugfix, before writing implementation code | Stars: 322k | Updated | Claude CodeCodex | Verified | Install |
| skill-creatoranthropics/skillsCreate new skills, modify and improve existing skills, and measure skill performance. Use when users want to create a skill from scratch, edit, or optimize an existing skill, run evals to test a skill, benchmark skill performance with variance analysis, or optimize a skill's description for better triggering accuracy. | Stars: 198k | Updated | Claude CodeCodex | Verified | Install |