Best fit
Finds LLM integration call sites that break when a model blocks a response or falls back to another model. Built for the Fable 5 era, where a high-risk call can be blocked and silently fall back to Opus 4.8. Read-only,…
// agent reliability
Guardrail Fallback Linter
Find the LLM calls that break when a model blocks a response or falls back to another model.
What it does
With Claude Fable 5, a call in a high-risk domain can be blocked and fall back to Claude Opus 4.8. Code that parses the response blind will break, or worse, quietly run on a downgraded answer nobody decided to accept. Guardrail Fallback Linter finds the call sites where that can happen.
It scans your source for LLM calls that assume every response is a usable answer: a call with no error or refusal branch, a response parsed with no guard, a hardcoded model id with no fallback handling. You get each finding so you can handle the block and the fallback on purpose instead of discovering it in production.
It reads files only. It does not run your code, call any model, or touch the network.
What it checks
- Calls with no error or refusal branch.
- Responses used or parsed with no guard.
- Hardcoded model ids with no fallback handling.
- Python and JavaScript or TypeScript call sites.
- An editable list of client call patterns you can extend.
- Each finding with a rule id, severity, file, line, message, and the evidence line.
What the output looks like
$ scan_guardrail_fallback.py src/
GFL001 HIGH client.py:42 response parsed, no refusal branch
GFL002 MED client.js:17 hardcoded model id, no fallback
read-only 2 findings, see remediation-snippets.md
One thing this is not
This is a reliability check. It helps you handle a guardrail block and a model fallback gracefully so your app degrades on purpose. It is not, and must not be used as, a way to bypass, weaken, or evade any safety guardrail.
// choose with context
Is Guardrail Fallback Linter the right skill?
It returns
Finds LLM integration call sites that break when a model blocks a response or falls back to another model. Built for the Fable 5 era, where a high-risk call can be blocked and silently fall back to Opus 4.8. Read-only,….
Do not use it as
Do not treat a successful run as proof beyond the supplied files, declared scope, and evidence the skill actually checked.
Compatibility, access, version, and licence
- Reads or accesses
- Only the files or context you deliberately provide; review the listing before granting additional access.
- Compatibility
- Agents that read the open SKILL.md format. Verify runtime and service requirements at the purchase or install destination.
- Version
- Verify the current package version at the linked destination.
- Licence
- Review the package licence and seller terms at the linked destination.
Questions
Does it run my code or call a model?
No. It reads your source files only. It does not run your code, call any model, or touch the network.
Which languages does it cover?
Python and JavaScript or TypeScript, using an editable list of common client call patterns. Detection is heuristic: it flags likely-unhandled call sites for review.
Is this a way around safety guardrails?
No. It is a reliability check that helps you handle a block and a fallback gracefully. It is not a way to bypass, weaken, or evade any safety guardrail.
Handle the block and the fallback on purpose.
Current price and purchase option are shown on Agensi. Instant download. Works with any agent that reads SKILL.md.
Get Guardrail Fallback Linter on Agensi