dataclaw

name: dataclaw description: > Export Claude Code, Codex, and other coding-agent conversation history to Hugging Face. Use when the user asks about exporting conversations, uploading to Hugging Face, configuring DataClaw, reviewing PII/secrets in exports, or managing their dataset. allowed-tools: Bash(dataclaw *), Bash(hf auth login ), Bash(pip install dataclaw), Bash(grep *)

DataClaw Skill

THE RULE

Every dataclaw command outputs next_steps. FOLLOW THEM.

Do not memorize the flow. Do not skip steps. Do not improvise. Run the command -> read the output -> follow next_steps. That's it.

Runtime guidance follows this checklist:

1. Install
2. Install skill
3. Prep 3A. Choose source scope 3B. Choose project scope 3C. Set redacted strings
4. Export locally
5. Review and confirm
6. Publish

The CLI tracks your stage as 1-4: auth -> configure -> review -> done. dataclaw export (push) is gated - you must run dataclaw confirm first or it will refuse.

Getting Started

Run dataclaw status (or dataclaw prep for full details) and follow the next_steps.

Output Format

• dataclaw prep, dataclaw config, dataclaw status, and dataclaw confirm output pure JSON
• dataclaw export outputs human-readable text followed by ---DATACLAW_JSON--- and a JSON block
• Always parse the JSON and act on next_steps

Key fields:

• stage / stage_number / total_stages - where you are
• next_steps - follow these in order
• next_command - the single most important command to run next (null if user input needed first)

PII Audit (Stage 5)

After dataclaw export --no-push, follow the next_steps in the JSON output. The flow is:

1. Ask the user their full name - then grep the export for it
2. Run the pii_commands from the JSON output and review results with the user
3. Ask the user what else to look for - company names, client names, private URLs, other people's names, custom domains
4. Deep manual scan - sample ~20 sessions (beginning, middle, end) and look for anything sensitive the regex missed
5. Fix and re-export if anything found: dataclaw config --redact "string" then dataclaw export --no-push
6. Run dataclaw confirm with text attestations - pass --full-name, --attest-full-name, --attest-sensitive, and --attest-manual-scan. It runs PII scan, verifies attestations, shows project breakdown, and unlocks pushing.
7. Push only after explicit user confirmation: dataclaw export --publish-attestation "User explicitly approved publishing to Hugging Face."

Commands Reference

dataclaw status                            # Show current stage and next steps
dataclaw prep                              # Discover projects, check HF auth
dataclaw prep --source all                 # Prep with all sources explicitly selected
dataclaw prep --source claude              # Prep using one supported source key (for example claude)
dataclaw confirm --full-name "NAME" --attest-full-name "..." --attest-sensitive "..." --attest-manual-scan "..." # Scan PII, verify attestations, unlock pushing
dataclaw confirm --file /path/to/file.jsonl --full-name "NAME" --attest-full-name "..." --attest-sensitive "..." --attest-manual-scan "..." # Confirm a specific export file
dataclaw list                              # List all projects with exclusion status
dataclaw list --source all                 # List all sources
dataclaw list --source claude              # List projects for one supported source key
dataclaw config                            # Show current config
dataclaw config --repo user/my-dataset     # Set HF repo
dataclaw config --source all               # REQUIRED source scope selection (<source|all>; for example claude or codex)
dataclaw config --exclude "a,b"            # Add excluded projects (appends)
dataclaw config --redact "str1,str2"       # Add strings to redact (appends)
dataclaw config --redact-usernames "u1,u2" # Add usernames to anonymize (appends)
dataclaw config --confirm-projects         # Mark project selection as confirmed
dataclaw export --publish-attestation "..." # Export and push (requires dataclaw confirm first)
dataclaw export --no-push                  # Export locally only
dataclaw export --source all --no-push     # Export all configured sources locally
dataclaw export --source claude --no-push  # Export one supported source scope locally
dataclaw export --all-projects             # Include everything (ignore exclusions)
dataclaw export --no-thinking              # Exclude extended thinking blocks
dataclaw export -o /path/to/file.jsonl     # Custom output path
dataclaw update-skill claude               # Install/update the dataclaw skill for Claude Code

Gotchas

• Never run bare hf auth login - it's interactive and will hang. Always use --token.
• --exclude, --redact, --redact-usernames APPEND - they never overwrite. Safe to call repeatedly.
• Source selection is REQUIRED before export - explicitly set dataclaw config --source <source|all> (for example claude or codex), or pass --source ... on export.
• dataclaw prep outputs pure JSON - parse it directly.
• Always export with --no-push first - review before publishing.
• dataclaw export (push) reuses the exact file reviewed by dataclaw confirm - if that file is missing, re-export locally and re-confirm before pushing.
• dataclaw export (push) requires dataclaw confirm first - it will refuse otherwise. Re-exporting with --no-push resets this.
• PII audit is critical - automated redaction is not foolproof.
• Large exports take time - 500+ sessions may take 1-3 minutes. Use a generous timeout.

Prerequisite

command -v dataclaw >/dev/null 2>&1 && echo "dataclaw: installed" || echo "NOT INSTALLED - run: pip install dataclaw"