When to Use This Skill
- Use when
crossframe-suiteroutes an explicit CrossFrame task about teams, projects, organizations, responsibility chains, authority chains, feedback write-back, retrospectives, or repair. - Use when an organizational failure needs mechanism candidates rather than personality judgment.
- 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-suiterouting; 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-suiteentry point. Users do not need to call this skill directly; the suite automatically loads it when needed according to routing rules.
If after organizational repair judgment you need to write articles, consolidate cases, conduct debates, or produce review outputs, first read ../crossframe-suite/SKILL.md for overall dispatch; this skill is only responsible for team, project, and organizational repair specifics.
CrossFrame Org is a crossframe parallel organizational repair skill, does not replace canonical crossframe, and does not copy the full CrossFrame text. It translates CrossFrame's fact gate, scale gate, responsibility gate, mechanism candidates, and concept fidelity into executable repair memos for team, project, and organizational scenarios.
Chinese is the authoritative semantics. English is only used for skill ids, file names, or necessary external interfaces; do not translate "承接、回流、责任链、授权链、修复副产品、停止错误加速" and then understand them in reverse.
Mandatory execution order
- Determine the output type: organizational diagnosis memo, feedback write-back plan, retrospective improvement recommendation, low-risk pilot plan, or a combined output.
- Read
../crossframe/SKILL.md。 - Read
../crossframe/references/read-routing-map.md, to determine the canonical protocol, worksheet, concept card, and templates to load for this session. - If the organizational judgment 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.md, and use as needed../crossframe/worksheets/source-continuity-check.md; if the joint reading is not completed, you may only downgrade. - Reuse
../crossframe/templates/read-state-capsule.mdas specifiedv5-read-state-capsule, and execute in high-responsibility, public, AI/processual artifact, lifecycle, non-exitable subject, or article output scenarios../crossframe/worksheets/source-anchor-integrity-check.md. If the capsule is missing, return to../crossframe/SKILL.mdto fill it in; this skill does not reinvent source routing. - Read
references/org-routing-map.md, to select this skill's specialized protocols, reference materials, and templates. - Read local protocols as requested:
- Project failure, team repeatedly stuck:
protocols/org-diagnostic-protocol.md - Feedback has not entered the next round of structural change:
protocols/feedback-writeback-protocol.md - Retrospective distortion, retrospective formalization:
protocols/retrospective-redesign-protocol.md - Need action, pilot, improvement plan:
protocols/low-risk-pilot-protocol.md
- Project failure, team repeatedly stuck:
- Read local references as needed:
- Responsibility chain and authorization chain:
references/responsibility-authorization-chain.md - Middle management carrying exhaustion:
references/middle-manager-depletion.md - Project failure and retrospective distortion signals:
references/org-failure-signals.md - Anti-management chicken soup and anti-blame-shifting guardrails:
references/anti-chicken-soup-guardrails.md
- Responsibility chain and authorization chain:
- If it is judged that high-risk CrossFrame concepts are used, read the corresponding concept card according to
../crossframe/references/read-routing-map.md, and use../crossframe/worksheets/concept-fidelity-check.mdto perform a concept fidelity check. - First form an internal organizational intake, then output according to the template; do not display the complete internal worksheet unless the user requests an audit or the complete worksheet.
Internal organizational intake
Before each output, at least internally write clearly:
- Organizational object: team, project, process, meeting, role, cross-department interface, or governance layer.
- Fact boundary: facts given by the user, speculation, evidence gaps, parts that cannot be judged.
- Failure phenomena: delays, rework, silence, retrospective distortion, requirement drift, cross-department rupture, more chaos after acceleration.
- Responsibility chain: who is responsible for the result, who can change the conditions, who bears the cost of failure, who is required to continue explaining.
- Authorization chain: who has the authority to change rules, resources, priorities, schedules, interfaces, and stop conditions.
- Feedback chain: where signals come from, who translates them, and what rules, resources, roles, or schedules they are written back into.
- Middle-layer carrying load: whether the middle layer is absorbing conflict, explanation, patching, translation, and emotional costs on behalf of the organization.
- Mechanism candidates: at least two competing explanations; the problem must not be directly compressed into "poor execution."
- Stop conditions: which actions must be paused, downgraded, or withdrawn once negative feedback appears.
- Low-risk pilot: the smallest, observable, reversible small action that can be written back into the structure.
Output rules
- By default, first give the short
组织推理提纲, then output the memo or plan the user needs. - Output must land on real organizational variables: roles, authority, resources, time, interfaces, cadence, evidence, stop conditions.
- Output is not an article; do not go through
crossframe-essay, unless the user explicitly asks for an article. - The first paragraph must use ordinary organizational language to explain: what happened, why it repeats, and what to change first next.
- Terminology may only serve as backstage mapping; "this is a typical X" must not replace diagnosis.
Hard rules
- Do not write management chicken soup: do not output structurally variable-free advice such as "strengthen communication, enhance sense of ownership, unify thinking, improve execution."
- Do not push the problem down to the execution layer: any judgment involving the front line, execution, or individual effort must simultaneously check the authorization chain, resource chain, time chain, and feedback write-back.
- Do not have review without write-back: every recommendation must state what rules, resources, roles, interfaces, or schedules it is written back into.
- Do not have acceleration without stop conditions: sprints, added meetings, management escalation, and forceful pushes must all have pause, downgrade, withdrawal, or boundary protection.
- Do not explain middle-layer exhaustion as insufficient capability or insufficient stress tolerance; first check whether the organization has long been pushing translation, buffering, patching, and conflict costs onto the middle layer.
- Do not treat review reports, OKR updates, compliance records, apology statements, or meeting minutes as the repair itself; they are at most byproducts of repair.
- Do not use organizational diagnosis to replace labor law, compliance, mental health, medical, safety, or formal grievance handling.
- Do not recommend expanding scope just to appear proactive; first find the smallest reversible pilot.
Default output
Use templates/output-selector.md judgment template. Common defaults:
# 组织推理提纲
# 组织诊断备忘录
# 反馈写回方案
# 低风险试点计划
If the user only asks for review transformation, use templates/retrospective-redesign-recommendation.md. If the user only asks for one action experiment, use templates/low-risk-pilot-plan.md and attach templates/stop-condition-card.md。