Skip to content
Back to blog

Introducing Agent Access — Fix User Feedback With Claude Code

Paste a feedback link into Claude Code. It reads the report and the screenshots, finds the code, and sends you a plan — then waits for your approval before it changes anything. Free on every plan.

October 2, 2026Vandeth Tho
feedbackclaude-codeaiagentsannouncement

The slow part of feedback isn't the fix

A user writes "the reference column shows a weird code" and attaches a screenshot. The fix might be ten lines. Getting there is the slow part: open the console, read the report, zoom into the screenshot, work out which screen it is, find the component, trace where the value comes from.

That is exactly the kind of work a coding agent is good at — as long as it doesn't go further than you want it to.

What Agent access does

Agent access lets Claude Code read your SupportDock feedback threads and propose fixes:

  1. Paste a feedback link into Claude Code.
  2. It reads the whole thread — message, app version, platform, URL, screenshots, PDF attachments, replies and your internal notes — then finds the code and investigates. Read-only.
  3. It sends you a plan. A summary of the report and exactly what it intends to change land on the thread as an internal note and in your Telegram chat.
  4. You approve or reject on the thread, with an optional note ("also cover the weekly view", "don't touch the migration"). Only on approval does it write the fix, verify it, and report back.

Guardrails, enforced by the server

An agent acting on a stranger's message needs limits that don't depend on a prompt:

  • Approval before action. The agent's only write is a proposal. It waits on your decision, and a rejection stops it.
  • It never talks to your users. An agent token can't reply, email the sender, archive a thread or change its status.
  • Your apps only, revocable. A personal token, stored hashed and shown once. It reads only your own feedback and stops working the moment you revoke it.

Set it up

  1. In the console, go to Settings → Agent access and create a token.
  2. Install the Claude Code skill and export the token:
mkdir -p ~/.claude/skills/supportdock-feedback && \
  curl -fsSL https://supportdock.io/skills/supportdock-feedback/SKILL.md \
  -o ~/.claude/skills/supportdock-feedback/SKILL.md
export SUPPORTDOCK_AGENT_TOKEN=sd_agent_your_token
  1. Paste any feedback link into Claude Code.

Using a different agent? It's three plain HTTP endpoints — read a thread, post a proposal, poll for your decision. See the Agent API docs.

Free on every plan

Agent access reads and proposes, so it costs nothing to run — it's included on Free and Pro. Learn more or create a token.

Introducing Agent Access — Fix User Feedback With Claude Code | SupportDock | SupportDock