About argument-selfloop
A workflow skill that creates a final argument snapshot from section drafts for paper-writing pipelines. It records structural labels, a consistency contract, and section hashes, then flags missing or undersized prose without making repairs.
Key features
- 01Summarize expected H3 sections
- 02Label paragraph argument moves
- 03Create a consistency contract
- 04Refresh section content hashes
- 05Flag missing or thin prose
Use cases
- 01Audit section drafts before merging
- 02Check paragraph-level argument structure
- 03Route failed sections for revision
Install
Copy the command and run it in a terminal.
npx skills add https:/ / github.com/ willoscar/ research-units-pipeline-skills --skill argument-selfloopAfter installing, type/argument-selfloopin Claude Code to confirm it loaded.
Bundled filesscripts/run.py
Details
From the project READMEArgument Snapshot
Triggers & routing
- Trigger: argument snapshot, section ledger, consistency contract, final section manifest, 论证快照, 一致性契约.
- Use when: section-level mutations are complete and the merge needs a current argument ledger plus content hashes.
Create the final C5 snapshot consumed by merge and later audit.
Despite the historical Skill name, the current implementation is not an autonomous rewrite loop. It reads the final H3 files, assigns bounded structural move labels, writes a compact consistency contract, refreshes section hashes, and fails when required prose is absent or too thin.
Position In The Workflow
section writing
-> style and numeric hygiene
-> logic polish
-> paragraph compaction
-> argument snapshot
-> optional transitions
-> merge
Running the snapshot after every section mutator prevents the ledger and manifest from describing an earlier draft.
Inputs
sections/outline/outline.ymlqueries.mdoutput/PARAGRAPH_CURATION_REPORT.md
Outputs
output/ARGUMENT_SELFLOOP_TODO.mdoutput/SECTION_ARGUMENT_SUMMARIES.jsonloutput/ARGUMENT_SKELETON.md- refreshed
sections/sections_manifest.jsonl
SECTION_ARGUMENT_SUMMARIES.jsonl contains one record per expected H3 and one
record per paragraph. Current move labels are deterministic signals drawn from:
setup, thesis, contrast, evidence, evaluation, limitation, synthesis, takeaway
They make section shape inspectable; they do not prove that an argument is
scientifically valid. ARGUMENT_SKELETON.md contains a
## Consistency Contract and a chapter-level map. It is a writer-facing audit
artifact, never reader-facing paper content.
PASS Contract
- every expected H3 exists;
- every H3 meets the profile paragraph floor;
- each paragraph has at least one allowed move;
ARGUMENT_SKELETON.mdcontains## Consistency Contract;sections_manifest.jsonlcarries currentbytesandsha256values.
On FAIL, inspect ARGUMENT_SELFLOOP_TODO.md and route the named section back to
its writer or evidence owner. This Skill does not apply the repair itself.
Run
uv run python .codex/skills/argument-selfloop/scripts/run.py \
--workspace workspaces/<name>
Optional runner fields are --unit-id, --inputs, --outputs, and
--checkpoint.