More screenshots

    Works with the AI Agents you already use

    Claude CodeClaude CodeCursorCursorCodex CLICodex CLIMuseMuseDotsDots+21 more

    πŸ“ Prompt Template Linter

    3

    The Prompt Template Linter is a specialized diagnostic tool designed to sanitize and optimize LLM prompt templates before they reach production.

    $12

    Secure checkout via Stripe

    Money-back guarantee

    Converts to your local currency at checkout

    Security scanned
    πŸ“ Prompt Template Linter

    πŸ“ Prompt Template Linter

    Example session with this skill installed

    Lint the prompt templates in the /prompts directory and check for injection risks and missing variables.

    • Read your context and instructions
    • Compiled the prompt template linter

    Findings

    • [PTL001] SEVERE: /prompts/user_gen.j2 (Line 12) - Untrusted variable 'user_input' interpolated into instruction region.
    • [PTL002] WARNING: /prompts/summary.md (Line 4) - Placeholder 'date_range' is defined but never used in the template.

    Connects securely to your tools. The creator never sees your data.

    About this skill

    What it does

    The Prompt Template Linter is a specialized diagnostic tool designed to sanitize and optimize LLM prompt templates before they reach production. It scans your template files to identify structural flaws, security risks, and logical inconsistencies that often lead to model hallucinations or prompt injection vulnerabilities.

    Why use this skill

    Manually auditing prompts for "jailbreak" surfaces or missing variables is error-prone. This skill automates the detection of critical issues like unescaped user input in instruction blocks, contradictory constraints, and orphaned placeholders. It ensures your prompts are robust, type-safe (where variables are concerned), and formatted for reliable parsing.

    Supported tools

    • File Formats: Supports .prompt, .txt, .md, .jinja, and .j2 files.
    • Integration: Works via CLI or stdin, making it ideal for pre-commit hooks or CI/CD pipelines.
    • Remediation: Includes a library of safe-snippet references to fix flagged issues instantly.

    Output format

    The skill produces a structured report categorized by rule ID (PTL001-PTL006) and severity, pinpointing the exact line and file where the issue resides.

    Heuristic linter. The injection check flags surface, not a confirmed exploit, contradiction detection is pair-based, and the variable check works from the placeholders it finds in the template. Treat findings as places to review.

    How to use

    1. 1

      Connect your agent to the Agensi MCP

      Copy the prompt and paste it into Claude, ChatGPT, Cursor or any agent that supports MCP. It sets up the connection for you. Manual setup

    2. 2

      Sign in to Agensi

      Your agent shows a sign-in link. Sign in and click Allow.

    3. 3

      Ask for what you need

      Your agent loads this skill and uses it on your task. It always asks before any payment.

    Reviews

    No reviews yet

    Be one of the first to try it. Every listed skill goes through the trust checks below.

    Security scanned

    Checked by our 8-point scan before listing

    Fresh listing

    Recently published to Agensi

    Money-back guarantee

    Doesn't do what its page promises? You get your money back.

    Trust & safety

    Security scan

    Security scan passed 5 months ago

    Permissions

    Read Files
    Terminal / Shell

    File Scopes

    **/*.prompt
    **/prompts/**/*.txt
    **/prompts/**/*.md
    **/*.jinja
    **/*.j2
    --stdin

    Read-only. It is a detector, not a prompt generator. Reads no environment variables and writes nothing.

    Listed5 months ago

    What's inside

    Frequently Asked Questions

    No, the Linter is a diagnostic tool designed to detect structural and security flaws. It flags errors like injection risks or contradictory instructions so you can fix them, but it does not generate or rewrite the prompt text for you.

    The skill is compatible with standard prompt file formats like .jinja, .md, and .txt, making it agent-agnostic. It can be integrated into any AI agent workflow, CI/CD pipeline, or pre-commit hook that uses local or cloud-based template files.

    The PTL001 rule specifically identifies "untrusted variable" flags where user input is placed directly into instruction blocks without delimiters or escaping. By flagging these injection surfaces, the tool helps you secure your prompts against unauthorized instruction overrides.

    Your purchase includes the core linting engine, a library of safe-snippet references for fixing flagged errors, and documentation for CLI/CI/CD integration. You will also receive lifetime access to future rule updates (PTL series) as new prompt vulnerabilities are discovered.

    Yes, the linter is particularly effective at finding "placeholder drift," such as variables defined in your code but missing from the template, or tags left in the template that aren't populated at runtime. This prevents the LLM from seeing confusing "filler" text that causes hallucinations.

    New in Agents & Orchestration