Skip to content

DexForce/EmbodiChain / .agents/skills/add-task-program

add-task-program

Verified

Create, compose, attach, validate, or repair an EmbodiChain Task Program. Use for program.yaml, integration.yaml, TaskProgramCfg, scene_binding, configured Task Program deployments, Semantic Call workflows, or Task Program expert demonstrations.

references/validation.md

View on GitHub
# Validation and Repair

Validate from inert configuration toward live execution. Report the highest
completed level and do not imply that a lower level proves physical success.

## Read-only static inspector

```bash
python .agents/skills/add-task-program/scripts/inspect_deployment.py \
  <task config>/task.<embodiment>.yaml
```

Pass multiple deployments to check embodiment variants in one run. Add
`--json --explain` for normalized declarations, authored paths, effective argument
contracts, resource selections and source bindings. Demo Skill deployments
require one environment and configured nominal sources require sequential flow.

The inspector:

1. loads the runnable config;
2. resolves environment and embodiment components;
3. composes the integration and execution policy;
4. checks semantic scene roots against native physical UIDs;
5. strictly loads the bound program;
6. runs catalog preflight and static Semantic Call lowering.

It never constructs a live simulator or registers a Gym environment.

### Resolved values and owners

Use `--explain` for a readable backend/component summary and composed
integration/policy parameters with their authored owners. Add `--json` for
`resolved_configuration`, typed runtime preset defaults, and machine-readable
source paths. The view distinguishes component values, inline/derived values,
deployment overrides, and built-in composite monitor SDK defaults. Other
monitor providers' defaults are not guessed. These values describe configuration,
not live measurements or evidence of physical success.

## Focused test surfaces

```bash
pytest -q tests/lab/task_program
pytest -q tests/gym/envs/task_program/test_configured_integration.py
pytest -q tests/gym/envs/task_program/test_task_vertical_slices.py
pytest -q tests/gym/envs/task_program/test_task_hand_over.py
pytest -q tests/test_task_program_package_data.py
```

Choose only tests relevant to the changed boundary. Add
`tests/gym/envs/task_program/` coverage for new configured service/provider
families.

## Repair order

Fix the earliest failing ownership layer:

1. syntax/closed fields;
2. missing component path;
3. duplicate component-owned inline fields;
4. absent physical `simulation_uid`;
5. scene/embodiment/policy contract mismatch;
6. missing logical resource, endpoint, capability, preset, or service;
7. unknown Semantic Call or missing registered lowerer;
8. compiler/preflight workflow conflict;
9. live grounding/execution/evidence failure.

Do not compensate for an earlier-layer failure with task-local runtime code.

## Physical qualification

After static validation, run the actual deployment with controlled
seeds/randomization. Check:

- grounding and planning success;
- controller-ready action shape;
- effect assurance and recovery;
- post-policies and validators;
- final row-local acceptance;
- persisted completion metadata.

Projected execution proves the intended command trajectory completed, not that
the physical task outcome was measured.
writing-skillsobra/superpowersUse when creating new skills, editing existing skills, or verifying skills work before deploymentStars: 319kUpdated Claude CodeCodexVerifiedInstall
xlsxanthropics/skillsUse this skill any time a spreadsheet file is the primary input or output. This means any task where the user wants to: open, read, edit, or fix an existing .xlsx, .xlsm, .xltx, .csv, or .tsv file (e.g., adding columns, computing formulas, formatting, charting, cleaning messy data); create a new spreadsheet from scratch or from other data sources; or convert between tabular file formats. Trigger especially when the user references a spreadsheet file by name or path — even casually (like "the xlsx in my downloads") — and wants something done to it or produced from it. Also trigger for cleaning or restructuring messy tabular data files (malformed rows, misplaced headers, junk data) into proper spreadsheets. The deliverable must be a spreadsheet file. Do NOT trigger when the primary deliverable is a Word document, HTML report, standalone Python script, database pipeline, or Google Sheets API integration, even if tabular data is involved.Stars: 196kUpdated Claude CodeCodexVerifiedInstall
shipping-and-launchaddyosmani/agent-skillsPrepares production launches. Use when preparing to deploy to production, or when asking what needs to be in place before shipping. Use when you need a pre-launch checklist, when setting up monitoring, when planning a staged rollout, or when you need a rollback strategy.Stars: 118kUpdated Claude CodeCodexVerifiedInstall
childhoodthedotmack/claude-memBabysit a shipped release after merge, publication, or deployment. Verify published packages, production behavior, installed plugins, and late CI or user reports until the observation window is clean; use babysit for PR readiness before release.Stars: 118kUpdated Claude CodeCodexVerifiedInstall