Glossary

Harness

Harness intercepts every Claude Code tool call before it runs. When a dangerous command matches a rule, the agent gets a plain-English explanation and steers around it — no hard crash, no lost work.

Rules Harness enforces

  • HIGH
    Recursive forced delete
    Blocks rm -rf on any path that is not a clearly scoped temp directory.
  • HIGH
    Credential exposure
    Prevents writing or echoing secret keys, tokens, and private certificates to stdout or files.
  • HIGH
    Force push to protected branches
    Intercepts git push --force targeting main or master before the command runs.
  • MEDIUM
    Cloud resource wipe
    Flags destructive cloud CLI commands (aws s3 rm --recursive, gcloud … delete) for review.
  • MEDIUM
    Global package clobber
    Warns before npm install -g or pip install --break-system-packages overwrite system tools.

How it fits the product

Harness is implemented as a PreToolUse hook in Claude Code. The hook fires on every bash command, file write, and edit before execution. When a tool call matches one of Harness's 50+ rules, the command is skipped and Claude Code receives a structured explanation — what matched, why it is dangerous, and what to do instead. The agent reads that explanation and continues with a safe alternative rather than crashing or waiting for human input. Rules are organised in three layers: system defaults shipped with Kontekstboard, enterprise or team policies synced to all members, and personal overrides each developer controls.

FeatureDetail
PreToolUse hookFires on every bash, write, and edit.
50+ rulesDestructive commands, credential exposure, force-push, cloud wipe.
Three layersSystem, Enterprise/Team, and Personal. All synced to Claude Code.
See Harness on the homepage