# Redline playbook

A reusable skill the "Redline the MSA before it hits your desk" recipe pulls in. Add it once and every agent on the team can use it.

Apply the attorney-authored playbook's fallback position to every recognized clause. Never substitute your own judgment of what is fair.

## Method
1. Load the redline playbook from memory before parsing the MSA.
2. Match each parsed clause to a playbook clause type (limitation of liability, indemnity, termination, IP assignment, and others in scope).
3. Pull that clause type's preferred position, fallback position, and walk-away line.
4. Compare the counterparty's language against the fallback position, not the preferred one.
5. Where the clause falls short of the fallback, draft the tracked-changes edit that reaches it, with a one-line rationale citing the playbook entry.
6. Where the clause type has no playbook entry, stop and do not draft an edit.

## Rules
- Never invent a fallback position; apply only what the playbook states.
- Unrecognized clause types are always flagged for {{reviewer}}, never redlined.
- Walk-away lines are reported for awareness only — only {{reviewer}} decides to walk.
- Clause types in scope for v1 vs. always-escalate (set on first run, then stored in memory).
- Counterparties on a do-not-auto-redline list are checked before drafting starts (set on first run, then stored in memory).

## Output
One tracked-changes edit per deviating clause, each tagged with clause type, playbook citation, and rationale; unrecognized clause types listed separately as flags, not edits.