Skip to content
Back to skills

Change Intake Compiler

ASecurity

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.

  • 7 stars
  • 0 votes
  • 0 copies
  • 5 views
  • Added September 8, 2026
securitygosecurity

Security analysis

A100/100

Pro scans all 6 files and shows the line behind each finding

Scanned September 22, 2026

npx -y skills add machenjie/rd-skills --skill change-intake-compiler --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Change Intake Compiler?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for Change Intake Compiler
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/machenjie-change-intake-compiler/badge)](https://www.skillsdirectory.com/skills/machenjie-change-intake-compiler)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

Download with Pro
SKILL.md
---
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 |

Files in this skill

  • SKILL.md4 KB
  • examples/example-output.md521 B
  • references/change-request-template.md1.6 KB
  • references/checklist.md521 B
  • references/evidence-patterns.md1.9 KB
  • references/index.md2 KB

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

View all in security →

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.