🛡️ Model Resilience Linter
Find the LLM integration code that will not survive a provider being pulled or going down. Flags single-provider lock-in with no alternative, calls with no failover branch, missing timeouts, retries with no limit or backoff, no degraded-mode default, and hardcoded endpoints with no alternate. This is about the model going away, not the model declining.
- Identify hardcoded LLM endpoints that lack secondary fallbacks.
- Detect missing timeouts and retry logic in AI service calls.
- Review repository for model provider lock-in and single points of failure.
$12
· or 60 creditsSecure checkout via Stripe
Included in download
- Identify hardcoded LLM endpoints that lack secondary fallbacks.
- Detect missing timeouts and retry logic in AI service calls.
- terminal, file_read automation included
- Ready for Cursor
Sample input
Check my backend directory to see if we have any single provider risks or missing failovers.
Sample output
### Model Resilience Audit Results **MRL001: Single-provider lock-in** - `services/gpt_client.ts:L42`: Direct dependency on OpenAI without fallback provider detected. **MRL003: No timeout configured** - `utils/llm_query.py:L12`: API call lacks an explicit timeout setting.
🛡️ Model Resilience Linter
Find the LLM integration code that will not survive a provider being pulled or going down. Flags single-provider lock-in with no alternative, calls with no failover branch, missing timeouts, retries with no limit or backoff, no degraded-mode default, and hardcoded endpoints with no alternate. This is about the model going away, not the model declining.
$12
· or 60 creditsSecure checkout via Stripe
Included in download
- Identify hardcoded LLM endpoints that lack secondary fallbacks.
- Detect missing timeouts and retry logic in AI service calls.
- terminal, file_read automation included
- Ready for Cursor
- Instant install
Sample input
Check my backend directory to see if we have any single provider risks or missing failovers.
Sample output
### Model Resilience Audit Results **MRL001: Single-provider lock-in** - `services/gpt_client.ts:L42`: Direct dependency on OpenAI without fallback provider detected. **MRL003: No timeout configured** - `utils/llm_query.py:L12`: API call lacks an explicit timeout setting.
About This Skill
Audit Your LLM Infrastructure for Single-Point-of-Failure Risk
In the rapidly shifting landscape of AI providers, relying on a single model or endpoint is a significant architectural risk. The Model Resilience Linter is a specialized static analysis tool designed for developers to identify brittle LLM integrations before a provider outage or API deprecation brings down your application.
What it does
The skill scans your codebase—supporting Python, JavaScript, and TypeScript—to detect patterns that signal poor reliability. It specifically looks for:
- Single-Provider Lock-in: Code that lacks a secondary fallback or failover logic.
- Configuration Gaps: Missing timeouts, hardcoded base URLs, and lack of exponential backoff.
- Graceful Degradation: Missing default states or "degraded-mode" handlers for when AI services are entirely unavailable.
Why use this skill?
Manually auditing a large repository for reliability anti-patterns is time-consuming and prone to human error. This skill automates the detection of infrastructure debt using a specialized rule set (MRL001-MRL006). Unlike general linters, it understands the specific failure modes of LLM calls, providing high-signal feedback and remediation references to help you build a provider-agnostic, resilient AI stack.
Use Cases
- Identify hardcoded LLM endpoints that lack secondary fallbacks.
- Detect missing timeouts and retry logic in AI service calls.
- Review repository for model provider lock-in and single points of failure.
- Ensure graceful degradation defaults are present for failed AI queries.
Known Limitations
Heuristic detector. It flags missing availability handling but cannot confirm that a failover actually works end to end. It is distinct from the Guardrail Fallback Linter, which handles a model declining a request rather than a provider disappearing.
How to Install
mkdir -p ~/.claude/skills && curl -sL https://www.agensi.io/api/install/model-resilience-linter -o /tmp/model-resilience-linter.zip && unzip -o /tmp/model-resilience-linter.zip -d ~/.claude/skills && rm /tmp/model-resilience-linter.zipFree skills install directly. Paid skills require purchase - use the download button above after buying.
Reviews
No reviews yet - be the first to share your experience.
Only users who have downloaded or purchased this skill can leave a review.
Early access skill
Be the first to review this skill.
Only users who have downloaded or purchased this skill can leave a review.
Security Scanned
Passed automated security review
Permissions
Allowed Hosts
File Scopes
It inspects the integration code around your LLM call sites. Reads no environment variables and writes nothing.
Works with any agent that can read a repo and run a local Python script (Claude Code, Cursor, Codex CLI, and other SKILL.md-compatible agents). Standard library only, no install step. Read-only, no network.