Skip to content

ThomasMoreAI/legal-skills-open / au/general/skills/triage-legal-enquiry-l0cka

triage-legal-enquiry-l0cka

Verified

Prepare a provisional, explainable and staff-reviewed triage record for a legal enquiry using an approved centre-specific Legal Triage profile. Use when authorised staff at an Australian community legal centre or legal assistance service need help identifying urgency, legal issue and jurisdiction, checking configured service pathways, documenting access needs, or preparing a referral. Do not use as a public chatbot, provide legal advice, receive unnecessary identifying information, perform conflict checks, assess credibility or merits, or make final eligibility, acceptance or rejection decisions.

FileKindSize
LICENSEother1.0 KB
SKILL.mdskill md5.3 KB
agents/openai.yamlconfig249 B
references/result-contract.mdmarkdown2.3 KB
references/triage-protocol.mdmarkdown4.8 KB

references/triage-protocol.md

View on GitHub
# Staff-facing triage protocol

Use this sequence for every enquiry. The approved centre profile controls where
this protocol refers to a configured rule or pathway.

## 1. Run the preflight

Confirm all of the following before processing substantive facts:

- an authorised staff member is operating the workflow;
- the centre profile passes `validate_triage_profile.py --require-approved`;
- the proposed AI environment and data class are approved by the centre; and
- the input contains no unnecessary identifiers or third-party information.

Do not solve a missing profile by relying on general legal knowledge. When an
item fails, do not solicit further substantive facts. If facts already supplied
indicate apparent urgent risk, return `URGENT HUMAN ESCALATION` and direct the
staff member to the centre's authorised urgent procedure. Otherwise return
`HUMAN TRIAGE REQUIRED`.

## 2. Check urgency first

Ask only enough to identify an immediate threat to safety or liberty, a current
detention or custody issue, an imminent court or tribunal event, a limitation or
filing deadline, threatened homelessness, or another trigger expressly listed
in the profile.

Use the matching configured action. Record the trigger and policy ID. Return
`URGENT HUMAN ESCALATION` and stop ordinary triage when the profile requires
immediate escalation.

Do not invent emergency contacts. If an apparent urgent risk has no configured
pathway, return `URGENT HUMAN ESCALATION` and direct the staff member to the
centre's authorised emergency procedure.

## 3. Preserve the conflict boundary

Perform conflict checking only through the centre's authorised process and
outside the model. Do not put names or identifying particulars into the prompt.

Record only:

- status: `passed`, `pending`, `uncertain` or `not started`;
- the authorised system or role that performed the check; and
- the time of the check, if supplied by staff.

If the check is not passed, gather no more substantive information than is
necessary for urgency and safe routing. Return `HUMAN TRIAGE REQUIRED` unless an
urgent escalation takes priority.

## 4. Gather minimum triage facts

Ask one short question at a time. Prefer the person's own description and
record uncertainty instead of translating it prematurely into a legal claim.
Gather only what the centre profile requires, ordinarily:

- the broad problem and the help sought;
- the relevant State, Territory, Commonwealth or other forum connection;
- key dates, documents already received and the next listed event;
- the type of other party without identifying that party; and
- communication, interpreter, disability, cultural-safety or other access needs
  voluntarily identified by the person.

Do not ask for traumatic detail merely to improve a classification. Do not
infer protected attributes, family violence, disability, capacity or financial
circumstances. Ask an approved question directly when the profile makes it
necessary and explain why it is being asked.

## 5. Compare with configured rules

Compare each relevant fact with the exact service area, jurisdiction,
eligibility factor or exclusion in the profile. For each conclusion, record:

- supplied fact;
- configured rule and source policy ID;
- provisional assessment;
- missing fact or assumption; and
- human decision required.

Do not assess legal merits, credibility, prospects, evidence quality or whether
the centre should form a lawyer-client relationship.

## 6. Select the status

Apply this priority order:

1. `URGENT HUMAN ESCALATION` for a configured or unresolved immediate risk.
2. `HUMAN TRIAGE REQUIRED` for a failed preflight, incomplete conflict check,
   profile gap, ambiguity, complexity or required professional judgment.
3. `INSUFFICIENT INFORMATION` when a bounded factual gap prevents routing.
4. `PROVISIONAL SERVICE PATHWAY` when a configured service may fit, subject to
   human confirmation.
5. `OUTSIDE CONFIGURED SCOPE` when the supplied facts clearly do not match the
   approved profile, subject to human confirmation and referral review.

Never substitute `eligible`, `ineligible`, `accepted`, `rejected` or a merits
rating.

## 7. Prepare a safe pathway or referral

For a possible internal pathway, name the service type and the next staff step.
Do not promise an appointment or acceptance.

For a referral:

- use only a current profile entry;
- explain the verified scope and any known access method;
- state that availability and acceptance remain with the receiving service;
- obtain consent before a warm referral or information transfer; and
- avoid transferring the triage narrative unless authorised and necessary.

## 8. Hand the draft to staff

Use the result contract. Label the record `DRAFT - HUMAN CONFIRMATION REQUIRED`.
Ask the authorised staff member to confirm, vary or reject the proposed pathway
before any client-system entry, advice, appointment, rejection or referral.
receiving-code-reviewobra/superpowersUse when receiving code review feedback, before implementing suggestions, especially if feedback seems unclear or technically questionable - requires technical rigor and verification, not performative agreement or blind implementationStars: 323kUpdated Claude CodeCodexVerifiedInstall
requesting-code-reviewobra/superpowersUse when completing tasks, implementing major features, or before merging to verify work meets requirementsStars: 323kUpdated Claude CodeCodexVerifiedInstall
using-git-worktreesobra/superpowersUse when starting feature work that needs isolation from current workspace or before executing implementation plans - ensures an isolated workspace exists via native tools or git worktree fallbackStars: 323kUpdated Claude CodeCodexVerifiedInstall
verification-before-completionobra/superpowersUse when about to claim work is complete, fixed, or passing, before committing or creating PRs - requires running verification commands and confirming output before making any success claims; evidence before assertions alwaysStars: 323kUpdated Claude CodeCodexVerifiedInstall