claude

llm-app-hardener

A repo-native toolchain for LLM apps: scaffolds eval suites, red-teams for prompt injection, hardens structured-output reliability, and audits token cost, riding promptfoo/DeepEval/Pydantic rather than replacing them

3 on GitHubby Sigistryv1.0.0MITupdated yesterdayai
/

Plugin llm-app-hardener from marketplace sigistry in sigistry/marketplace. Commands read from the repo's manifest — yesterday.

This plugin runs code on your machine — its hooks fire automatically on events like tool use. The exact commands are listed below; read them before installing.

What it installs

Skills (4)

  • eval-authoring

    This skill should be used when the user mentions "llm eval", "evaluation", "promptfoo", "deepeval", "regression test", "llm judge", "golden dataset", "eval suite", "test a prompt", or wants to prove a prompt/model change improved rather than regressed behavior. It provides a standardized methodology

  • owasp-llm-top10

    This skill should be used when the user mentions "prompt injection", "jailbreak", "owasp llm", "llm security", "guardrails", "data exfiltration", "insecure output", "tool abuse", "indirect injection", or "is this agent safe". It provides the OWASP Top 10 for LLM Applications mapped to code-level sig

  • prompt-versioning-hygiene

    This skill should be used when the user mentions "prompt engineering", "prompt versioning", "prompt management", "system prompt", "prompt template", "manage prompts", or "prompts scattered in code". It provides a methodology for treating prompts as versioned, reviewable, testable code rather than ma

  • structured-output

    This skill should be used when the user mentions "structured output", "json mode", "tool schema", "function calling", "json parse error", "response_format", "pydantic ai", "invalid json from the model", or "the model's JSON keeps breaking". It provides a methodology for getting reliably valid, schem

Slash commands (4)

  • /llm-app-hardener:eval-scaffold

    Scaffold a real eval suite for an app's prompts/agents so a change can be proven better, not eyeballed

  • /llm-app-hardener:rag-tune

    Audit a RAG pipeline for retrieval quality and cost, and emit prioritized concrete fixes

  • /llm-app-hardener:structured-output-doctor

    Find LLM calls whose output is parsed as JSON and harden them with native structured output plus validate-and-retry

  • /llm-app-hardener:token-cost-audit

    Audit LLM call sites for token and cost waste, quantify the savings, and apply the safe wins

Subagents (2)

  • prompt-injection-auditor

    Use this agent when you need to red-team an LLM app's prompt construction and tool-calling for OWASP LLM Top 10 risks, especially LLM01 prompt injection: untrusted input (user text, retrieved documents, tool outputs, web/email content) concatenated into prompts without delimiting or guardrails, tool

  • token-cost-optimizer

    Use this agent when you need to apply token and cost optimizations to LLM call sites, enabling prompt caching for stable prefixes, trimming redundant or re-sent context, routing clearly-easy tasks to a cheaper model tier, setting sensible max_tokens, and batching independent calls, while explaining

Hooks (1)

  • PostToolUse

    node "${CLAUDE_PLUGIN_ROOT}/hooks/prompt-detect.mjs"

View source on GitHub sigistry.com

More from sigistry/marketplace

All 30 plugins →

Monitoring and evaluation jobs on your project folder: twelve job skills (evaluation questions, question-method alignment, evaluation TOR drafts, indicator reference sheets, measurement feasibility, M&E plan drafts, progress evidence, findings to actions, results chain alignment, sampling plans, data quality assessment plans, baseline planning), three workflow packs and nineteen commands that cross-reference logframes, M&E plans, proposals, reports and TORs and write editable Markdown with source traceability.

24 skills · 19 commands

3

me-review

claude

Twelve structured monitoring and evaluation document reviews: logframes, theories of change, indicators, evaluation TORs, M&E plans, proposals, sampling plans, surveys, baseline reports, progress reports, data quality assessments, and evaluation reports. Each returns section-by-section PASS/PARTIAL/FAIL ratings, a summary verdict, and prioritized recommendations.

36 skills

3

ballast

claudecodex

Builds up to a goal in a field you have no expertise in: splits it into the foundations that goal actually needs, mobilizes what you already hold, learns what is missing, and leaves every solved path behind for the next goal. Nothing is called done until a check passes.

16 skills · hooks

3

Triages failing GitHub Actions runs, hardens Dockerfiles, pre-flights Kubernetes manifests before they reach the cluster, and drafts blameless postmortems from incident evidence

4 skills · 5 commands · 3 agents · hooks

3

Goes past detecting accessibility and localization issues to fixing them: cross-element ARIA and keyboard-focus remediation, color-contrast rewrites, hardcoded-string extraction, ICU MessageFormat correctness, and RTL readiness

3 skills · 5 commands · 2 agents · hooks

3

Owns everything around the diff: craft Conventional Commits and PR descriptions, recommend the right semver bump by reasoning about behavioral breaking changes, assemble changelogs from messy history, and guide merge-conflict resolution, rebase planning, and git-history archaeology

2 skills · 5 commands · 3 agents · hooks

3

Similar plugins

AI-powered codebase understanding — analyze, visualize, and explain any project

9 skills · 10 agents · hooks

Egonex-AI/Understand-Anything 92,410

career-ops

claudecodex

Open-source AI job search that runs locally in your AI coding CLI: scan job portals, evaluate listings into a structured A-H report with a 1-5 score, tailor your CV, track applications

5 skills · hooks

career-ops-hq/career-ops 82,253

humanizer

claude

Rewrite AI-sounding text so it reads like the writer without changing what it says.

1 skill

blader/humanizer 68,461

mempalace

claudecodex

AI memory system — mine projects and conversations into a searchable palace. 47 MCP tools, auto-save hooks, guided setup.

3 skills · 6 commands · 1 MCP server · hooks

MemPalace/mempalace 60,536

Perform AI code review on Git diffs — supports workspace changes, branch ranges, and single commits with concurrent per-file analysis, codebase search, and deep context-aware review.

2 commands

alibaba/open-code-review 50,068

llm-finetuning

claudecodex

Eval-gated LLM fine-tuning lifecycle (LoRA/QLoRA, DPO, GRPO/RLVR, vision SFT, quantized export). Use when fine-tuning an open-weights model end-to-end: building the eval harness first, selecting method and model, preparing data, training with Unsloth or TRL, gating checkpoints, and exporting GGUF/FP8.

10 skills · 2 commands · 3 agents

wshobson/agents 42,461

Maintain llm-app-hardener?

This page updates itself from your repo. Add the badge to your README, or get it Featured to pin it above the organic results with a followed link to your site.

claude & codex plugins30 plugins