skills.rest
VoxTechnologies
VoxTechnologiesOFFICIAL
Explore all skills by this author →
WebsiteGitHub

feedback

Processes user feedback batches through spec updates, code changes, tests, and documentation.

4|Updated Mar 14, 2026
Search Tags:#feedback loop#continuous improvement#triage#user feedback#spec updates#agent teams#brd prd
One-click install
npx skills add https://github.com/VoxTechnologies/transmute-framework --skill feedback-voxtechnologies
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: feedback
Source: https://github.com/VoxTechnologies/transmute-framework/tree/main/skills/feedback
Command: npx skills add https://github.com/VoxTechnologies/transmute-framework --skill feedback-voxtechnologies

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve? Post-launch products accumulate user feedback from support tickets, analytics, and stakeholders, but acting on it consistently is hard: specs drift from code, conflicting requests go unresolved, and documentation falls behind. This Skill runs a structured feedback loop that triages each item, updates BRD/PRD specifications, implements code changes, and keeps tests and docs in sync. ## Core Features & Use Cases - Feedback Triage and Conflict Resolution: Categorizes items (Bug, UX Issue, Missing Feature, Enhancement, Performance, Documentation), traces them to existing BRD/PRD requirements, detects conflicting requests, and assigns APPROVED / STAKEHOLDER_DECISION_REQUIRED / DEFERRED statuses. - Sequential Agent Team Execution: Coordinates three teammates — spec-updater, code-implementer, and test-and-docs-updater — with enforced dependency ordering and a maximum of two coordination rounds per item. - Living Document Maintenance: Updates BRD/PRD with traceability markers, refreshes tests, developer docs, and Mintlify user guides, then archives each processed batch for recurring-request detection. - Use Case: After collecting a month of support tickets in feedback/input.md, run Stage 8 to triage 15 items, update the affected specs, ship the fixes on a dedicated branch, and produce a resolution report with a full audit trail. ## Quick Start Ask the agent to process the user feedback batch in feedback/input.md and run the Stage 8 feedback loop.

Frequently Asked Questions about feedback

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I process user feedback into code changes automatically?▼

Create a structured feedback/input.md file with categorized items (ID, category, description, affected feature), then run the Stage 8 feedback loop. It triages items, updates BRD/PRD specs, implements code changes, and updates tests and docs through three sequential teammates.

What format should the feedback input file use?▼

Each feedback item needs an ID (e.g., TICKET-001), a category (Bug, UX Issue, Missing Feature, Enhancement, Performance, or Documentation), a description, and an affected feature ID (FEAT-XXX or NEW). Optional fields include frequency, user quote, and source.

How are conflicting feedback requests resolved?▼

Items are grouped by affected feature and conflicts are classified by type: quantitative conflicts resolve by frequency, simplicity-versus-power may use feature flags, and safety-versus-convenience escalates to stakeholders. Unresolvable conflicts are marked STAKEHOLDER_DECISION_REQUIRED with a 48-hour deadline.

Can Stage 8 run at the same time as dependency maintenance?▼

No. Stage 8 and Stage 9 (Dependency Maintenance) must never run concurrently because both modify package.json and lock files. Verify no chore/dependency-update branch is active before starting the feedback loop.

What happens if the feedback stage is interrupted mid-execution?▼

Recovery checks for marker files: analysis.md means triage finished, change-plan.md means planning finished, and resolution.md means the stage completed. Partial spec updates without code changes must be completed before ending the session to keep specs and code in sync.

How many feedback items can one batch process?▼

The default limit is 15 APPROVED items per run, configurable via the feedback batch limit in tech-stack.md. Batches estimated over 3-5 hours of effort are split into priority-ordered sub-batches, with the remainder deferred to the next cycle.

Related Skills

Explore More Product & Management Skills→
omeraltnomeraltn

agency-feedback-synthesizer

Synthesize user feedback into prioritized product insights with sentiment scores and RICE rankings.

Community
Advanced
ldmrepoldmrepo

regression-check

Classify residual regression risk from diffs and verification reports.

Community
Advanced
virajpvirajp
1

feedback

Classify production feedback into bugs, metrics, UX issues, and feature requests.

Community
Advanced
PatriotAiPatriotAi
5

surgical-code-refactoring

Patches large existing codebases with dependency-aware edits and pre-integration validation.

Community
Advanced