docs: add anonymizer Claude Code skill and supporting concept docs

[lipikaramaswamy]

Summary

• Claude Code skill at skills/anonymizer/ — elicits dataset context, recommends mode + strategy, drafts a runnable Python script, iterates with the user via the workflow in interactive.md.
• docs/concepts/choosing-a-strategy.md (215L) — decision guide for mode (Replace vs Rewrite), strategy choice, privacy goal phrasing, and detection knobs. Doubles as the primary agent reference.
• docs/troubleshooting.md (210L) — symptom-first guide for dropped rows, leakage, low utility, and pipeline failures.
• README.md — new "Using with Claude Code" section pointing at the skills.sh installer.
• mkdocs.yml — adds the two new docs to navigation.
• src/anonymizer/__init__.py — exports PrivacyGoal at the top level (previously only accessible via deep import despite being referenced in docs and the skill's output template).
• docs/concepts/detection.md — minor wording polish.

Testing notes

Tested end-to-end with 20 PMC-Patients clinical case reports. The interactive workflow was exercised: install verify → provider check → data inspect → clarify → plan → build → preview → results inspection. Two real workflow gaps surfaced and fixed during testing:

• Provider configuration wasn't checked proactively → added step 1 environment verification + step 3 provider question (commit 8ccd277)
• Trace data wasn't persisted by default → script now saves preview.parquet for inspection (commit 5c5d5a0)

The skill loader was also verified: npx skills add (skills.sh) discovers the skill correctly via the SKILL.md frontmatter.

Related issues filed during this work

• chore: use COL_* constants for intermediate columns in llm_replace_workflow.py #148 — replace string-literal column names in llm_replace_workflow.py:53-55 with COL_* constants (existing STYLEGUIDE rule violation)
• bug: result.trace_dataframe is not persistable via to_parquet #152 — result.trace_dataframe not persistable via to_parquet. Blocks the skill's preview.parquet save from being readable until fixed. Same 3 lines as chore: use COL_* constants for intermediate columns in llm_replace_workflow.py #148.

Test plan

• CI passes (format-check, copyright-check, mkdocs build --strict)
• from anonymizer import PrivacyGoal succeeds
• Skill discoverable via npx skills add against this branch
• Skill workflow produces a sensible script for a sample dataset (verified manually with PMC-Patients data)

[greptile-apps]

Greptile Summary

This PR adds a Claude Code skill at skills/anonymizer/ along with two new concept documents (docs/concepts/choosing-a-strategy.md, docs/troubleshooting.md), a README section, and a PrivacyGoal top-level export in src/anonymizer/__init__.py. The skill elicits dataset context, guides mode/strategy selection, and produces a runnable Python script following an interactive workflow.

• New docs: choosing-a-strategy.md is a full decision guide covering mode, strategy, detection knobs, and privacy-goal phrasing; troubleshooting.md is a symptom-first guide for dropped rows, leakage, low utility, and pipeline failures. Both are wired into mkdocs.yml navigation.
• PrivacyGoal export: Previously accessible only via anonymizer.config.rewrite, it is now exported at anonymizer.__init__ and properly listed in __all__.
• Known tracked issue: result.trace_dataframe.to_parquet(\"preview.parquet\") in the skill's output template will raise a serialization error until issue bug: result.trace_dataframe is not persistable via to_parquet #152 is resolved — this is acknowledged in the PR description and has already been flagged in a previous review comment.

Confidence Score: 4/5

Safe to merge for docs and the PrivacyGoal export; the skill preview path has a known crash on every first run that is tracked but not yet fixed.

The PrivacyGoal export, all doc content, and the interactive workflow are solid. The one live defect is in the skill template: result.trace_dataframe.to_parquet raises an unhandled exception before the failure-first guard and quality summary can execute on every preview run.

skills/anonymizer/SKILL.md — the output template preview path crashes on trace_dataframe serialization until issue #152 is fixed.

Important Files Changed

Filename	Overview
skills/anonymizer/SKILL.md	New skill definition with output template; the preview path calls result.trace_dataframe.to_parquet() which is non-serializable per issue #152, causing a crash before the failure-first guard runs (already flagged in previous review).
src/anonymizer/init.py	Adds PrivacyGoal re-export from anonymizer.config.rewrite; import path is correct, all updated correctly.
docs/concepts/choosing-a-strategy.md	New decision guide; leakage/tolerance values match code constants, DEFAULT_ENTITY_LABELS unpacking syntax is correct, all cross-doc links resolve properly.
docs/troubleshooting.md	New symptom-first guide; leakage threshold table matches RiskToleranceBundle values in config/rewrite.py exactly, relative links are correct.
skills/anonymizer/workflows/interactive.md	Interactive workflow steps are consistent with the API; Anonymizer(model_providers=...) parameter verified against the constructor signature.
AGENTS.md	Adds a pointer to the new skill and a note to check SKILL.md before shipping public-API changes.
README.md	Adds a Using with Claude Code section with install command; straightforward documentation addition.
mkdocs.yml	Adds two new nav entries for choosing-a-strategy and troubleshooting; both files exist and are in the right directories.
docs/concepts/detection.md	Minor wording change; no logic change.

_{Reviews (8): Last reviewed commit: "docs: mention bundled agent skill in AGE..." | Re-trigger Greptile}

[lipikaramaswamy]

Once #149 merges, this PR needs a follow-up edit to AGENTS.md:

1. The dev-vs-user redirect at the top should also mention the bundled skill at skills/anonymizer/
2. A note that public-API changes (especially imports referenced in skills/anonymizer/SKILL.md's output template) may require corresponding skill updates.

[asteier2026]

mrn is in the default list, as is employee-id and possibly case_number and court_name

[lipikaramaswamy]

Good catch — fixed in de0ab26. Cross-checked all examples against the actual DEFAULT_ENTITY_LABELS list and dropped redundant ones (mrn, court_name, employee_id). Also switched all examples to snake_case to match the convention — the validator only strips/lowercases, so medical record number and medical_record_number would have been treated as different labels.

Note on case_number: verified it's not in DEFAULT_ENTITY_LABELS, so I kept that one

[alexahaushalter]

Are dropped rows and failed rows different? My first assumption is they are the same, and then it is confusing to me to make sure there were no dropped rows but to then investigate any that failed.

[lipikaramaswamy]

Good catch — they were the same thing. Unified in 0c46cc0: the intro now uses one term ("rows that didn't make it through the pipeline, usually a rate-limit / infra issue") and explicitly contrasts with quality-issue rows (high leakage_mass, low utility_score, needs_human_review).

[greptile-apps]

Want your agent to iterate on Greptile's feedback? Try greploops.