Add Document Type Definitions (DTD) support for job outputs (#81)

* Add Document Type Definitions (DTD) support for job outputs

Implement DTDs as a first-class feature for formalizing document specifications
with quality criteria, enabling consistent document structure across job outputs.

Key changes:
- Add DTD schema and parser for frontmatter markdown format
- Extend job.yml output schema to support DTD references
- Update skill generators to inject DTD context into templates
- Add document detection workflow to deepwork_jobs.define
- Add DTD improvement workflow to deepwork_jobs.learn
- Update Claude and Gemini templates for DTD-aware rendering
- Bump version to 0.4.0

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

* Rename dtd to document_type with full file path references

Change the output schema from referencing DTDs by name to using full file paths:
- Renamed `dtd` field to `document_type` in job.yml outputs
- Changed pattern from name-based (e.g., "monthly_report") to path-based
  (e.g., ".deepwork/dtds/monthly_report.md")
- This makes references self-documenting and allows agents to understand
  them without additional context

Updated:
- Schema: new pattern ^\.deepwork/dtds/[a-z][a-z0-9_-]*\.md$
- Parser: OutputSpec.dtd -> OutputSpec.document_type
- Generator: load DTDs by file path instead of directory scan
- Templates: has_dtd -> has_document_type, dtd -> document_type
- Documentation: updated all examples and explanations
- Tests and fixtures: renamed job_with_dtd to job_with_document_type

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

* Add job.yml DTD for define step output validation

- Create job_spec.md DTD with 9 quality criteria for job.yml files
- Update install.py to copy DTDs from standard jobs to .deepwork/dtds/
- Reference DTD in deepwork_jobs define step output with document_type
- Bump deepwork_jobs version to 0.7.0
- Temporarily disable new-standard-job-warning rule (bug causes infinite loop)

The DTD ensures consistent quality for all job.yml files created via
/deepwork_jobs.define by validating: identifier format, semantic version,
summary length, description richness, changelog presence, step completeness,
dependency validity, input consistency, and output paths.

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

* Add review_job_spec step to deepwork_jobs workflow

Add DTD-based quality validation step between define and implement that:
- Uses a sub-agent (haiku model) for unbiased review of job.yml
- Evaluates all 9 DTD quality criteria from job_spec.md
- Iterates fix-review cycle until all criteria pass

This establishes a reusable pattern for DTD-conforming document validation.

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

* Add outputs to process-oriented steps for DTD compliance

Review of all jobs against job_spec DTD revealed 3 jobs with steps
missing required outputs:
- add_platform/verify: added verification_checklist.md
- commit (test, lint, commit_and_push): added implicit state outputs
- update/job: added files_synced implicit state output

These are process-oriented steps where the "output" is a state change
rather than a file artifact, but the DTD requires at least one output
per step for completeness.

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

* Rename DTD to doc spec throughout codebase

- Rename dtd_parser.py to doc_spec_parser.py
- Rename dtd_schema.py to doc_spec_schema.py
- Rename .deepwork/dtds/ directories to doc_specs/
- Rename document-type-definitions.md to doc-specs.md
- Update all references from DTD/dtd to doc spec/doc_spec
- Update schema pattern for document_type paths
- Update documentation, job definitions, and templates
- Add pytest and uv to allowed commands in Claude settings

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

---------

Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>

Author: nhorton

.claude/settings.json

@@ -89,6 +89,8 @@
       "Bash(python3:*)",
       "Bash(pip:*)",
       "Bash(pip3:*)",
+      "Bash(pytest:*)",
+      "Bash(uv:*)",
       "Bash(node:*)",
       "Bash(npm:*)",
       "Bash(npx:*)",

.claude/skills/add_platform.add_capabilities/SKILL.md

@@ -208,7 +208,9 @@ Use branch format: `deepwork/add_platform-[instance]-YYYYMMDD`
 ## Outputs
 
 **Required outputs**:
-- `job_schema.py`- `adapters.py`
+- `job_schema.py`
+- `adapters.py`
+
 ## Quality Validation
 
 Stop hooks will automatically validate your work. The loop continues until all criteria pass.

.claude/skills/add_platform.implement/SKILL.md

@@ -306,7 +306,10 @@ Use branch format: `deepwork/add_platform-[instance]-YYYYMMDD`
 ## Outputs
 
 **Required outputs**:
-- `templates/` (directory)- `tests/` (directory)- `README.md`
+- `templates/` (directory)
+- `tests/` (directory)
+- `README.md`
+
 ## Quality Validation
 
 Stop hooks will automatically validate your work. The loop continues until all criteria pass.

.claude/skills/add_platform.research/SKILL.md

@@ -258,7 +258,9 @@ Use branch format: `deepwork/add_platform-[instance]-YYYYMMDD`
 ## Outputs
 
 **Required outputs**:
-- `cli_configuration.md`- `hooks_system.md`
+- `cli_configuration.md`
+- `hooks_system.md`
+
 ## Quality Validation
 
 Stop hooks will automatically validate your work. The loop continues until all criteria pass.

.claude/skills/add_platform.verify/SKILL.md

@@ -182,7 +182,8 @@ Use branch format: `deepwork/add_platform-[instance]-YYYYMMDD`
 
 ## Outputs
 
-No specific file outputs required.
+**Required outputs**:
+- `verification_checklist.md`
 
 ## Quality Validation
 
@@ -195,7 +196,7 @@ Stop hooks will automatically validate your work. The loop continues until all c
 ## On Completion
 
 1. Verify outputs are created
-2. Inform user: "Step 4/4 complete"
+2. Inform user: "Step 4/4 complete, outputs: verification_checklist.md"
 3. **Workflow complete**: All steps finished. Consider creating a PR to merge the work branch.
 
 ---

.claude/skills/commit.commit_and_push/SKILL.md

@@ -133,7 +133,8 @@ Use branch format: `deepwork/commit-[instance]-YYYYMMDD`
 
 ## Outputs
 
-No specific file outputs required.
+**Required outputs**:
+- `changes_committed`
 
 ## Quality Validation
 
@@ -146,7 +147,7 @@ Stop hooks will automatically validate your work. The loop continues until all c
 ## On Completion
 
 1. Verify outputs are created
-2. Inform user: "Step 3/3 complete"
+2. Inform user: "Step 3/3 complete, outputs: changes_committed"
 3. **Workflow complete**: All steps finished. Consider creating a PR to merge the work branch.
 
 ---

.claude/skills/commit.lint/SKILL.md

@@ -128,7 +128,8 @@ Use branch format: `deepwork/commit-[instance]-YYYYMMDD`
 
 ## Outputs
 
-No specific file outputs required.
+**Required outputs**:
+- `code_formatted`
 
 ## Quality Validation
 
@@ -141,7 +142,7 @@ Stop hooks will automatically validate your work. The loop continues until all c
 ## On Completion
 
 1. Verify outputs are created
-2. Inform user: "Step 2/3 complete"
+2. Inform user: "Step 2/3 complete, outputs: code_formatted"
 3. **Continue workflow**: Use Skill tool to invoke `/commit.commit_and_push`
 
 ---

.claude/skills/commit.test/SKILL.md

@@ -111,7 +111,8 @@ Use branch format: `deepwork/commit-[instance]-YYYYMMDD`
 
 ## Outputs
 
-No specific file outputs required.
+**Required outputs**:
+- `tests_passing`
 
 ## Quality Validation
 
@@ -124,7 +125,7 @@ Stop hooks will automatically validate your work. The loop continues until all c
 ## On Completion
 
 1. Verify outputs are created
-2. Inform user: "Step 1/3 complete"
+2. Inform user: "Step 1/3 complete, outputs: tests_passing"
 3. **Continue workflow**: Use Skill tool to invoke `/commit.lint`
 
 ---