Use `analysis-agent` for ambiguous or missing intent: slice desired behavior, boundaries, constraints, and completion signals. Skip an accepted Brief or repository-independent answers fully determined by user facts.
Installs into .claude/skills of the current project.
Sponsors
SSkills DirectorySkills Directory
Is that skill safe to install?
SSkills DirectorySkills Directory
Know which skills are safe — weekly.
Are you the author of Change Intake Compiler?
Add the live security badge to your README. It updates with every re-scan.
[](https://www.skillsdirectory.com/skills/machenjie-change-intake-compiler)
---
name: change-intake-compiler
description: "Use `analysis-agent` for ambiguous or missing intent: slice desired behavior, boundaries, constraints, and completion signals. Skip an accepted Brief or repository-independent answers fully determined by user facts."
---
# change-intake-compiler
## Role
Support `analysis-agent` in separating source-backed change intent from
reversible assumptions and user-owned decisions.
## When To Use
- ambiguous request
- missing desired behavior
## Do Not Use
- accepted Engineering Brief already exists
- questions fully determined by user-supplied facts that require no repository evidence or engineering intent compilation
## Required Inputs
- request summary
- observed current behavior
## Professional Decision Rules
- Define the intent slice as the smallest independently observable unit containing current behavior, desired behavior, acceptance or completion signal, constraints, non-goals, and affected surfaces.
- Separate facts discoverable from source from choices only the user can make.
- Proceed with explicit reversible assumptions only when they cannot change public behavior, data, authorization, or domain meaning.
- Reject solution wording that hides the actual problem or success condition.
## High-Value Gotchas
- Do not turn an ambiguous request into an invented requirement.
- A source-discoverable fact is not a reason to interrupt the user.
- Make every reversible assumption explicit in the output.
## Gap Classification
- Treat a discoverable fact as research work: inspect the authorized source and record the evidence instead of asking the requester to supply repository facts.
- Treat a reversible assumption as provisional only when it is explicit, bounded, and cannot alter public behavior, data, authorization, or domain meaning.
- Treat a user-owned decision as blocking when engineering evidence cannot determine product semantics, scope, priority, or another stakeholder-owned tradeoff.
- Treat an unsafe or non-reversible unknown as an unresolved requirement; do not hide it inside an implementation assumption.
## Execution Checklist
1. Reconstruct current behavior, desired behavior, acceptance or completion signal, constraints, non-goals, and affected surfaces from the request and source evidence.
2. Classify each gap as discoverable fact, reversible assumption, or user-owned choice.
3. Choose the smallest intent slice that retains every required field and whose acceptance can be observed independently.
4. Stop routing when a missing choice can change contract, data, authority, or domain meaning.
## Stop / Escalation Conditions
- Stop planning when current behavior, desired behavior, constraints, non-goals,
completion signal, authority, or affected surface is missing.
- Escalate assumptions that could change contract, data, security, migration,
rollback, user-visible behavior, or acceptance.
- Stop solution-first requests when the named implementation is not backed by an outcome, constraint, authority, and rejected simpler option.
- Stop conflict resolution when stakeholders disagree on behavior, priority, rollout, role, date, or acceptance and no decision owner or deadline is named.
- Escalate when required source channels, issue trackers, production evidence, or stakeholder context cannot be inspected through authorized, redacted, fresh read/search access.
## Output Contract
- minimal intent slice with current behavior, desired behavior, observable acceptance or completion signal, constraints, non-goals, and affected surfaces
- assumptions
- open questions
## Targeted References
| Path | Type | Load when | Do not load when | Required by | Required output |
|---|---|---|---|---|---|
| [change request](references/change-request-template.md) | template | Drafting or normalizing the durable Change Request artifact | Only readiness classification or gap triage is needed | analysis-agent | completed-contract |
| [checklist](references/checklist.md) | decision-checklist | Auditing intake readiness quickly before routing downstream | The request is blocked by a single missing authority or evidence fact | analysis-agent | checklist-result, residual-risk |
| [evidence patterns](references/evidence-patterns.md) | evidence-pattern | Closure depends on raw-input-to-field mapping, stakeholder authority, source freshness, blocking-question evidence, or proof limits | No evidence conflict, authority, or freshness issue is material | analysis-agent | evidence-record, proof-limit, residual-risk |
| [index](references/index.md) | index | competing change intake compiler references require dependency, conflict, or output-fragment selection | the change intake compiler root or a task-named reference already resolves selection | analysis-agent | reference-selection |