GitGuardian SaaS
Sources Blog / feed → AI Hooks keep leaked tool output away from the model (2026-09-25) Blog Rank Why this rank top features scored 90/80/60, +10 product media, +6 runnable example — 92.67 How ranking works →GitGuardian SaaS is a cloud-based security platform that detects and prevents secrets leakage in git repositories and source code.
GitGuardian SaaS made ggshield's API timeout configurable to stop large-repo scans from failing, and expanded AI Hooks to withhold secrets from model context on more coding assistants while extending coverage to new agent CLIs.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄
what's the number?
Each feature carries 0–100 for how completely the vendor documented it — not how big or important the work is. A major capability described in eight words scores low, and that is the finding.
- depth0–40 what it does and how it works or what changed
- specificity0–30 names real surfaces — APIs, flags, formats, limits, numbers
- actionability0–30 enough to go use it — a named endpoint, flag, or config key tops this; a UI path is a starting point
Code and config rank first by construction: a documented endpoint, flag or config key scores at the top of specificity and actionability, so it sorts above a feature described only in prose. Under 60 the notes go thin — everything below the “thinner coverage” line is thinner documentation, not smaller work. Hover any meter for that feature's three sub-scores.
ggshield previously enforced a fixed 60-second API timeout that could cause large-repo scans to fail with read timeouts.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ A new api_timeout config key (set in ~/.gitguardian/.gitguardian.yaml) and a GITGUARDIAN_API_TIMEOUT environment variable let users raise this limit, e.g.︀︄︇︀︅︉︀︅︀︀︃︀︀︅︅︀︅︊︀︄︆︀︃︄ GITGUARDIAN_API_TIMEOUT=120 ggshield secret scan repo . or api_timeout: 120 in the config file.
GITGUARDIAN_API_TIMEOUT=120 ggshield secret scan repo .api_timeout: 120
api_timeout: 120
GITGUARDIAN_API_TIMEOUT=120 ggshield secret scan repo .AI Hooks and the AI agents/MCPMCPModel Context Protocol, a standard for connecting AI applications to external tools and data sources. A compatible server exposes capabilities an assistant can use through a common interface. inventory now cover Amazon Kiro (IDE and CLI) and Junie CLI, blocking secrets in commands and file reads. A new ggshield machine setup subcommand registers AI Hooks for Kiro and Junie CLI in one step.
$ ggshield machine setup
AI Hooks now withhold shell-command output containing secrets from the model context on Claude Code, Codex, and Mistral Vibe — the assistant receives only the ggshield message, keeping the secret out of the model session transcript.