ワンクリックで
playbook-lookup
Look up the firm's playbook position on a named clause type and return the preferred / acceptable / non-negotiable framing.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
メニュー
Look up the firm's playbook position on a named clause type and return the preferred / acceptable / non-negotiable framing.
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
SOC 職業分類に基づく
| slug | playbook-lookup |
| name | Playbook Lookup |
| description | Look up the firm's playbook position on a named clause type and return the preferred / acceptable / non-negotiable framing. |
| version | 0.1.0 |
| metadata | {"sources":[{"mode":"port-original","author":"ClaudeCoder","added_in":"0.1.0"}]} |
Port-original skill. Hand-authored for Legal & Contracts. Used by drafters and reviewers.
A drafting or review skill needs the firm's stated position on a specific clause type — e.g., "what's our position on liability cap in MSAs?" "what's our default NDA term?"
Required: contract type (NDA / MSA / SOW / offer / vendor / other) + clause type (e.g., "liability cap", "term", "indemnity scope", "non-compete"). Optional: client / matter context (the playbook may carry per-client overrides).
A structured lookup result:
clause_type: <e.g., liability-cap>
contract_type: <e.g., MSA>
playbook_position:
preferred: <wording or summary>
acceptable: <fallback wording or range>
non_negotiable: <hard floor / red line>
rationale: <one-line why>
notes:
- <provenance: when added, by whom, last updated>
- <client-override if any>
exists: true | false
If exists: false, return that explicitly so the caller knows to fall back to conservative defaults and enqueue clause-library-curate to fill the gap.
skills/clause-library-curate maintains it).exists: false and propose adding one.Lookup-only. This skill never interprets the playbook position or applies it to a specific draft — that's the caller's job.
Per-client overrides must be sourced (which matter, which attorney approved); silent overrides are a red flag and should be flagged to escalation-routing.
UPL line. This skill is designed against the unauthorized-practice-of-law line. Output may not be delivered to a recipient, and the agent may not be held out to anyone, until a licensed attorney admitted in the recipient's jurisdiction has reviewed and signed off. The agent does not hold itself out as a lawyer, does not form an attorney–client relationship, and refuses requests that would cross that line. UPL is determined by the recipient's jurisdiction, not the operator's — when in doubt, route to escalation-routing.
Return the result inline as YAML or attach it to the work-item thread for the caller to consume.
Draft a board (or shareholder) consent for a routine action.
Apply a triggering event to the cap table and emit the diff.
Summarize a case (or set of cases) on a named question.
Add a newly observed clause to the clause library with its classification, source, and the playbook position it relates to.
Sequence handoffs when legal work spans multiple teams.
Draft offer letters and employment agreements from a hiring brief.