no token tax
No token tax: why Quell ships zero mandatory AI
If your team already pays for a coding agent, why should its feedback tool sell you another pile of tokens? The case for a deterministic feedback rail and flat seats.
The Quell Blog
Field notes on agent-native feedback.
Notes from building a feedback product around coding agents: what worked, what was only a neat claim, and which boring design decisions mattered more than the model.
Cornerstone · Show HN
We built a feedback widget that lets a coding agent fix its own bug reports
Most bug reports discard the evidence an agent needs, then blame the agent for guessing. This is the shape of a better handoff: the report, the runtime context and the resulting change.
$ quell context 412
## Bug — export button no-op
console TypeError: r.blob is not a fn
network 500 POST /api/export
page /reports?range=q4
→ agent: reproduced · PR #412 opened
status: review-required
More posts
trust-aware routing
Why feedback trust levels matter for safe agent automation
The dangerous sentence in agent software is "it can do this automatically". Trust levels answer the missing question: on whose authority?
Claude Code
How to give Claude Code access to user feedback
Claude Code can search your codebase. It still cannot see the bug report sitting in another inbox. Give it evidence through a read-only MCP connection first.
feedback MCP
MCP server for user feedback: what your agent needs
The MCP connection is the easy part. What the agent gets when it calls get_context is what determines whether the connection is useful.
agent-readable feedback
What to include in a bug report
What belongs in a bug report changes when a coding agent reads it. The six fields that matter for agent consumption, and why prose descriptions alone produce 'cannot reproduce'.
vibe coding
The missing feedback loop in vibe-coded apps
AI builders have made shipping a v1 a matter of hours. The harder problem starts the day after: knowing what broke. Here is what the Observe step looks like.
Claude Code
Closing the feedback loop in Claude Code and Cursor
Claude Code and Cursor handle the coding half. The feedback half is unsolved: knowing what broke after users touch the app. Here is how to close it.
trust-aware routing
How to automate bug report triage without losing control
Bug report volume grows when you ship with AI. The answer is separating reports that already carry agent-level authority from reports that need a human to read them first.
AI coding agents
Feedback tools for AI coding agents: the category that search results miss
Search for a feedback tool that works with AI coding agents and you find code review tools. The category you actually need — user feedback to agent — is almost unclaimed.
Claude Code
Feedback tool for Claude Code: what teams actually need
Most feedback tools were built with a human on the other end. Four things separate tools built for Claude Code teams from tools that added a webhook and called it agent-ready.
agent-readable feedback
What your coding agent needs to fix a user-reported bug
A user writes one sentence. A coding agent needs console logs, network calls, and page state. Here is how to close the gap without asking the user to open DevTools.
vibe coding
How to add user feedback to a Lovable or Bolt app
Lovable and Bolt make shipping a v1 fast. Add one script tag and your coding agent can read the console trace, network errors, and page URL attached to every bug report.
Comparisons