crossframe-casebook

CrossFrame Casebook

When to Use This Skill

• Use when crossframe-suite routes explicit CrossFrame materials into reusable casebook entries, anonymized case records, mechanism extraction, or retrieval indexes.
• Use when the goal is future reuse rather than immediate advice.
• Do not use independently unless the user explicitly names this sibling skill.

Packaged Source Note

This AAS-ready copy preserves the original CrossFrame skill body below. Chinese remains the canonical semantic layer; English metadata is only for discovery, installation, and repository review.

Limitations

• The skill body is intentionally Chinese-canonical; English metadata is for discovery and does not replace the original Chinese terms.
• Use only after explicit CrossFrame invocation or crossframe-suite routing; do not apply it as a generic default reasoning layer.
• It structures analysis, drafting, and review, but does not replace source verification, domain expertise, or legal, medical, or financial judgment.

This skill does not trigger independently. All CrossFrame tasks are uniformly dispatched from the crossframe-suite entry. Users do not need to call this skill directly; the suite loads it automatically when needed according to routing rules.

If, after case distillation, there is still a need for write-up, teaching, debate, or public/organizational special judgment, first read ../crossframe-suite/SKILL.md for overall dispatch; this skill is only responsible for case library entries and reusable material structures.

CrossFrame Casebook is a parallel case library skill to crossframe and does not replace crossframe. It is only responsible for organizing material into reusable case entries: first hold the boundaries of facts, sources, and privacy, then extract scale windows, mechanism chains, responsibility chains, reverse conditions, reusable concepts, and follow-up observations.

Chinese is the authoritative semantics. English is only used for skill id, file names, field names, or external introductions; when Chinese and English conflict, the Chinese terminology prevails.

Sequence that must be executed

1. Read ../crossframe/SKILL.mdand confirm the basic CrossFrame gates and expression boundaries that this material should follow.
2. Read ../crossframe/references/read-routing-map.mdand, according to the material topic, select the CrossFrame protocol, concept cards, and judgment tiers that need to be aligned.
3. If the material triggers high responsibility, public institutions, intimate relationships, long-term evolution, framework governance, AI reality verification, weak signals/opacity, inability to exit, instrumentalization, metaphor/source transparency, or article output, you must additionally read ../crossframe/references/continuity-bundles.mdand use ../crossframe/worksheets/source-continuity-check.mdas needed; if the joint reading is not completed, you may only downgrade.
4. Reuse ../crossframe/templates/read-state-capsule.md as stipulated in v5-read-state-capsule, and execute ../crossframe/worksheets/source-anchor-integrity-check.mdin high-responsibility, public, AI/procedural artifact, lifecycle, non-exitable subject, or article output scenarios. If the capsule is missing, return to ../crossframe/SKILL.md to complete it; this skill does not reinvent source routing.
5. Read protocols/material-boundary-protocol.mdand first stratify sources, facts, speculation, privacy, and publishability.
6. Read protocols/casebook-build-protocol.mdand decide whether this session is to create a new case, clean old cases, batch-index, compare cases, or turn a retrospective into a case library.
7. Read references/casebook-field-guide.mdand ensure that each case distills at least nine items: case summary, fact boundaries, material sources, scale windows, mechanism chains, responsibility chains, reverse conditions, reusable concepts, and follow-up observations.
8. Read references/privacy-and-redaction-rules.mdand de-identify individuals, organizations, place names, times, original chat text, screenshots, links, and identifiable details.
9. Read protocols/mechanism-extraction-protocol.mdand extract mechanism chains and responsibility chains from story narration, avoiding writing only plot or piling up concepts.
10. Read templates according to the task: for a single case read templates/casebook-entry-template.md; for batch cases read templates/casebook-index-template.md; when source audit is needed read templates/redacted-source-ledger-template.md。
11. Before output, do a smoke check: do not treat guesses as facts, do not leak privacy, do not write only stories without extracting mechanisms, and do not pile up concepts.

Input handling

• Chat records: retain interaction structure, role relationships, observable behavior, and chronological order; delete or generalize names, accounts, contact information, precise locations, and irrelevant private details.
• Organizational material: distinguish formal institutions, oral conventions, meeting minutes, project records, personal feelings, and secondhand accounts.
• Project retrospective: distinguish result facts, process facts, explanations, responsibility attribution, remedial actions, and unverified assumptions.
• Public controversy: distinguish public sources, statements by involved parties, media reports, platform rules, legal facts, public-opinion interpretations, and model speculation; when the latest facts or real people or organizations are involved, sources must be checked.

Default output

By default, output one or more 案例库条目. Each entry must contain at least:

• Case summary
• Factual boundaries
• Material sources
• Scale window
• Mechanism chain
• Responsibility chain
• Reverse conditions
• Reusable concepts
• Follow-up observations

If the user requests that a case library be maintained, then additionally add 案例索引、标签、相似案例、复用场景 and 更新记录。

Hard rules

• Do not copy the full text of crossframe ; read the canonical skill and routing map only through relative paths.
• Do not directly preserve chat transcripts or personal information as case assets unless the user explicitly requests it and the publicly shareable scope has been confirmed.
• Do not write speculation, motive inferences, or secondhand evaluations as facts.
• Do not just tell stories; each case must extract at least one mechanism chain and one responsibility chain.
• Do not use CrossFrame terminology to replace case facts; concepts must serve reuse, not decorate the output.
• Do not write the case library as a personality judgment, organizational conviction, public-opinion verdict, or compliance endorsement.
• Do not use public scale to erase personal harm, organizational dereliction, evidence gaps, or responsibility chains.
• When evidence is insufficient but the risk is urgent, only low-risk, retractable, observable follow-up observation items may be given.

Quality gate

A qualified casebook output must be able to answer:

• What structural problem can this case be reused to identify?
• Which materials are facts, and which are only interpretations or speculation?
• Are this case's sources traceable, de-identifiable, and publicly shareable?
• Which scale window is currently being used, and has improper scale shifting occurred?
• How does the mechanism chain move from conditions, behavior, and feedback to outcomes?
• In the responsibility chain, who has the power to change the conditions, and who bears the cost?
• What reverse conditions would overturn or downgrade this case's judgment?
• Which concepts truly improve reusability, and which are merely piles of terminology?
• When similar materials are encountered next time, what signals should be observed?