@umbrella-coop/flare-redact-claude-code
v0.1.0
Published
Claude Code plugin: redact secrets/PII via PreToolUse/PostToolUse/UserPromptSubmit hooks
Readme
@umbrella-coop/flare-redact-claude-code
Claude Code plugin that redacts secrets and PII before they reach Claude or leak
from tool results, powered by flare-redact.
Install
# local directory (from the monorepo)
claude plugin install ./pkg/claude-code/plugin
# or via a marketplace
/plugin marketplace add umbrella-coop/flare-redact-ai-code-assistant
/plugin install flare-redact@umbrella-coop-flare-redact
/reload-pluginsValidate before shipping: claude plugin validate ./pkg/claude-code/plugin --strict.
The hook script is a single self-contained bundle — no install step, no
node_modules in the plugin.
What it guards
| Hook | Behaviour |
|---|---|
| PreToolUse | Deny write tools / shell commands carrying secrets; allow with redacted args for read-like tools |
| PostToolUse | Replace tool output with a masked, shape-preserving copy before Claude sees it |
| UserPromptSubmit | Block prompts containing secrets (Claude cannot rewrite prompts) |
Slash commands: /flare-redact:scan, /flare-redact:mask, /flare-redact:status
(model invocation disabled — user-invoked only).
Configuration
Put a flare-redact.config.json in the project root. See the
configuration reference.
Optional env vars:
FLARE_REDACT_AUDIT_FILE— append value-free JSONL audit events to a file.
Security
Findings and audit events never contain raw secret values. Writes and sensitive
reads (*.env, *.pem, secrets/**, ...) are blocked by default. Detection is
best-effort — hooks are a guardrail, not a complete enforcement boundary.
License
MIT
