Skip to content
Kent Kent
2026 Respawn OSS

Bash Hooks

Overview

Section titled “Overview”

Kent can post-process shell command output before it is shown to the model. It normalizes terminal output, reduces command noise, and adds useful execution context. Kent keeps this separate from command execution itself:

  • command runs normally
  • Kent runs the configured post-processing chain
  • successful commands omit the exit-code header; failed commands include Exit code N, output:
  • The model can disable that when it needs the full output

Config

Section titled “Config”

Configure command post-processing under [shell] in ~/.kent/config.toml:

[shell]
postprocessing_mode = "all" # none | builtin | user | all
postprocess_hook = "~/.kent/shell_postprocess_hook"

postprocessing_mode

Section titled “postprocessing_mode”

Allowed values:

  • none: disable command post-processing
  • builtin: run Kent’s sanitizer and built-in processors
  • user: run Kent’s sanitizer, then your configured hook
  • all: run Kent’s sanitizer, built-ins, then your configured hook

raw: true on a shell tool call bypasses command post-processing for that call. raw: true and postprocessing_mode = "none" preserve ANSI/style sequences in command output, subject to JSON result encoding and output truncation.

Built-ins

Section titled “Built-ins”
  • Successful direct go test ... commands collapse to PASS when output has no benchmarks, coverage, or JSON data.
  • Direct partial file reads add information about file size to the output, like [Total line count: 186]
  • Built-in processors run as a chain. A processor can pass output to later processors or stop the built-in chain with final output.

Hook Protocol

Section titled “Hook Protocol”

Input

Section titled “Input”

Kent sends JSON like:

{
  "tool_name": "exec_command",
  "command": "go test ./...",
  "parsed_args": ["go", "test", "./..."],
  "command_name": "go",
  "workdir": "/abs/workdir",
  "original_output": "...sanitized command output...",
  "current_output": "...built-in processed output or original output...",
  "exit_code": 0,
  "backgrounded": false,
  "max_display_chars": 16000
}

Your hook receives both:

  • original_output: sanitized command output before Kent semantic shaping
  • current_output: current Kent output after built-ins, or the same as original_output if no built-in handled it

This lets your hook either add on top of Kent defaults or replace them completely. In all mode, the hook runs after built-ins even when a built-in processor stops the built-in chain.

Hook must return JSON like:

{
  "processed": true,
  "replaced_output": "...new output..."
}

Return {"processed": false} for no-op passthrough. If the hook is missing, times out, exits nonzero, or returns invalid JSON, Kent falls back to the current output and reports a warning.

Edit page