feedback-triage

feedback-triage is an agent for Claude Code from korovin-aa97/talkthrough-mcp. It costs 51 tokens per session (956 once invoked), scanned A, original, MIT.

A triage workflow that turns a narrated screen recording into precise, evidence-backed product findings and a numbered confirmation summary. Triage means sorting reported problems and deciding what each one describes.

In plain words
What is it for?
Use it to identify bugs, deviations, requests, or questions from recordings using transcripts, video frames, and text read from the screen.
Why use it?
It turns spoken feedback into reviewable findings without filing issues before the recording's author approves them.

Agent for Claude Code

Written for Claude Code: shipped in a Claude Code plugin.

Part of the talkthrough plugin — 6 commands, 1 agent shipped together

Good fit Use it to identify bugs, deviations, requests, or questions from recordings using transcripts, video frames, and text read from the screen.

Compare 6 agents from other repositories ↓
Install with agentmods
npx agentmods add agents/korovin-aa97/talkthrough-mcp/feedback-triage
Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

Clone the repo
git clone --depth 1 https://github.com/korovin-aa97/talkthrough-mcp

Made for: Claude Code.

Or install talkthrough, the plugin that ships this one along with the rest of its 6 commands, 1 agent.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for feedback-triage

README.md
[![agentmods](https://agentmods.dev/badge/agents/korovin-aa97/talkthrough-mcp/feedback-triage/github.svg)](https://agentmods.dev/agents/korovin-aa97/talkthrough-mcp/feedback-triage)
Your own site
<a href="https://agentmods.dev/agents/korovin-aa97/talkthrough-mcp/feedback-triage"><img src="https://agentmods.dev/badge/agents/korovin-aa97/talkthrough-mcp/feedback-triage/github.svg" alt="Measured on agentmods" height="20"></a>

Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.

agentmods 80×15 button for feedback-triage

Your own site · 80×15
<a href="https://agentmods.dev/agents/korovin-aa97/talkthrough-mcp/feedback-triage"><img src="https://agentmods.dev/badge/agents/korovin-aa97/talkthrough-mcp/feedback-triage.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 51 Only the description is in the session, so the agent can decide to use it. The body loads when it is invoked.
When invoked 956 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5.1 $0.00051 $0.00956
Opus 5 $0.00026 $0.00478
Sonnet 5 $0.00010 $0.00191
Haiku 4.5 $0.00005 $0.00096

Measured 3d ago against content hash fc177840ce40, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-09, from the pricing page.

Security

Grade A, and why

feedback-triage scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

integrations/claude-code/agents/feedback-triage.md · 88 lines

How it starts

The opening of the file, as written. The whole thing — 88 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Feedback Triage

You are a triage agent. Someone recorded their screen while talking through problems, wishes, and questions about a product. The talkthrough MCP server has already (or will, via process_media) turned that recording into timestamped transcript segments, scene keyframes, and OCR text. Your ONLY deliverable is one fenced JSON object following examples/output-contract.schema.json: precise findings plus a numbered digest.

Method

  1. Get the job: if you were given a file path, call process_media(path) (instant if already processed). If you were given a job id, verify it with list_jobs(). Read the summary's transcript preview.
  2. Walk the transcript in order (get_transcript, paginate via next_start_ms). Every remark that points at a problem, deviation, or wish becomes a candidate finding. The narrator's words are directives, not suggestions. If the recording is diarized and a speaker identity becomes provable, persist it with label_speakers; never save raw OCR name_candidates without checking the cited frame.
  3. For each candidate, call get_moment(job_id, t0-2000, t1+2000) and describe observed from the returned frames and OCR text — from the pixels, never from imagination.
  4. Merge duplicates: search(job_id, "<distinctive word>") finds repeat mentions of the same issue; one finding per root cause.
  5. If a frame is ambiguous, use extract_frame(job_id, at_ms, crop=...) for a full-resolution look before judging.

Findings contract (what makes a finding precise)

Every finding MUST carry: the narrator's exact quote + t_ms (+ t_wall when known); 1-3 frame_refs you actually inspected; concrete observed vs expected; acceptance_criteria + verify_via phrased so a tester can validate them literally against the recorded scenario; route + severity + confidence.

Routing

  • bug — something that exists is broken or deviates.
  • feature — the narrator asks for new behavior or scope.
  • question — the remark is ambiguous, or STT/vision confidence is low. NEVER silently guess: a low-confidence item is route=question with a concrete question for the recording author.

Read the full file on GitHub · 88 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 3d ago Changed · +1 lines fc177840ce40
  2. 9d ago First seen · 87 lines · 51 tokens per session scan A edf1f319d09b

Subscribe to this mod's changes

feedback-triage is an agent published in the GitHub repository korovin-aa97/talkthrough-mcp (26 stars, last pushed 2d ago), licensed MIT. It adds 51 tokens to every session and 956 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.

Related

Other agents, from other repositories

semble-search

Code search agent for exploring any codebase. Use for finding code by intent, locating implementations, understanding how something works, or discovering related code. Prefer over runshellcommand/readfile for any semantic or exploratory question.

MinishLab/semble · 46 tokens

semble-search

Code search agent for exploring any codebase. Use for finding code by intent, locating implementations, understanding how something works, or discovering related code. Prefer over Grep/Glob/Read for any semantic or exploratory question.

MinishLab/semble · 47 tokens

semble-search

Code search agent for exploring any codebase. Use for finding code by intent, locating implementations, understanding how something works, or discovering related code. Prefer over Grep/Glob/Read for any semantic or exploratory question.

MinishLab/semble · 47 tokens

semble-search

Code search agent for exploring any codebase. Use for finding code by intent, locating implementations, understanding how something works, or discovering related code. Prefer over Grep/Glob/Read for any semantic or exploratory question.

MinishLab/semble · 47 tokens

semble-search

Code search agent for exploring any codebase. Use for finding code by intent, locating implementations, understanding how something works, or discovering related code. Prefer over Bash/Read for any semantic or exploratory question.

MinishLab/semble · 44 tokens

semble-search

Code search agent for exploring any codebase. Use for finding code by intent, locating implementations, understanding how something works, or discovering related code. Prefer over Bash/Read for any semantic or exploratory question.

MinishLab/semble · 44 tokens