SKILLEMALL.ai
GitHub ↗ RU Sign in

DC detecting-indirect-prompt-injection

Detect and defend against indirect prompt injection hidden in web pages, documents, and images consumed by an agent, via content extraction (HTML/PDF/OCR), normalization, and scanning with LLM Guard's PromptInjection scanner or Hugging Face Prompt Guard 2. Use when an agent ingests untrusted external content and you need to screen it for injected instructions before the LLM processes it.

Not recommendedcritical or high security findings · low grade D
mukul975/Anthropic-Cybersecurity-Skills Agent Skills author: mukul975 Apache-2.0 5 files · 1 script body ≈ 2 578 tokens Open the sourcegithub.com↗ analyzed 5 d ago

Detect and defend against indirect prompt injection hidden in web pages, documents, and images consumed by an agent, via content extraction (HTML/PDF/OCR)…

As a process C 56/100 · Has gaps — weak spots: result and completion, when it triggers, failures and branches

AnalyzerAI and agentsSecuritytype and topics are labelled automatically from the skill text
Runs in: Claude / Claude Code ChatGPT / Codex Gemini CLI Cursor GitHub Copilot / VS Code Qwen Code Kimi Code CLI Z.ai (GLM) DeepSeek (Deep Code) Mistral Vibe LangChain (deepagents) OpenClaw Goose OpenCode
JSON
Technical rating
D
76/100
safety, quality, tests
Safety 60%
77
Quality 40%
75
Run on models
none yet
Process rating
C
56/100
Has gaps
Result and completion w 14
0
Failures and branches w 10
0
When it triggers w 12
20
the three weakest of ten parameters · all ten

What is at stake

The skill contains fragments that, in the wrong hands, cost money or data. Below: what the installer risks and what the author should do.

Obfuscation
If you install

The files contain invisible characters, encoded commands or comments hidden from readers but visible to the model. What you read differs from what the agent sees.

For the author

Remove invisible characters (they usually sneak in through copy-paste) and encoded strings: no catalog will pass them. Instructions for the model must be readable by a human too.

How to improve

  1. Address the high-severity findings: each costs 18 safety points. If one is a false positive, add the rule id to guard.allow in spec.yaml.
For the model run — optional
  • Your own cases (evals/evals.json, 4–6 real requests with expected answers): the full check would then run those instead of a model-drafted suite.
  • A spec.yaml with trigger phrases and assertions — a behaviour contract for CI; `skilltest init` writes a template.

Guard findings · 2

  • high Obfuscation uni-zero-width SKILL.md:115
    Zero-width / invisible characters (possible hidden text) (5 occurrences)
    ZERO_WIDTH = dict.fromkeys(map(ord, "␀␀␀␀␀"), None)
Medium and low: 1
  • medium Obfuscation uni-zero-width scripts/agent.py:24
    Zero-width / invisible characters (possible hidden text) (5 occurrences)
    ZERO_WIDTH = dict.fromkeys(map(ord, "␀␀␀␀␀"), None)

Files scanned: 5. Evidence is masked. Grey chips explain why severity was lowered.

Against the Agent Skills spec

  • note frontmatter-key unknown frontmatter key "domain"
  • note frontmatter-key unknown frontmatter key "subdomain"
  • note frontmatter-key unknown frontmatter key "nist_ai_rmf"
  • note frontmatter-key unknown frontmatter key "atlas_techniques"

Process rating: all ten parameters 56/100

  • 0Result and completion. Does not say what the result is
  • 0Failures and branches. Linear process with no failure handling
  • 20When it triggers. No condition that starts the skill
  • 60Tools and files. Uses tools (python) that frontmatter does not declare
  • 70Inputs and preconditions. Inputs and preconditions are listed
  • 100Steps. 23 steps
  • 100Consistency. Name and required fields are in place
  • 100Execution cost. Instruction body is 2578 tokens
  • 100Running it twice. No mutating operations
  • 100Progress reporting. Reports progress

Everything here is measured from the skill text rather than judged by a model, so the numbers are checkable. A parameter weighs more when it is a more common reason for the process to stall.

Quality signals

  • +5Description has no quoted example phrases that should trigger the skill
  • +4Description does not say when NOT to use the skill (false activations)
  • +3Output format is not stated: the model decides each time
  • -42 reference files, but SKILL.md never points to them: the model will not open them
  • -31 of 1 scripts are never mentioned in SKILL.md
  • +2Single-language instructions
  • +3Description length 390: enough signal without eating the budget
  • +4Structure: 17 headings
  • +3Step-by-step instructions: 23 items
  • +4Has examples (7 code blocks)
  • +1License stated

Quality base 70; lint remarks subtract, signals add up to 100. Result: 75.