Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx skills add EresusSecurity/appsec-skills --skill eresus-deser-auditgit clone --depth 1 https://github.com/EresusSecurity/appsec-skillsWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/eresussecurity/appsec-skills/eresus-deser-audit)<a href="https://agentmods.dev/skills/eresussecurity/appsec-skills/eresus-deser-audit"><img src="https://agentmods.dev/badge/skills/eresussecurity/appsec-skills/eresus-deser-audit/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/eresussecurity/appsec-skills/eresus-deser-audit"><img src="https://agentmods.dev/badge/skills/eresussecurity/appsec-skills/eresus-deser-audit.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00084 | $0.02038 |
| Opus 5 | $0.00042 | $0.01019 |
| Sonnet 5 | $0.00017 | $0.00408 |
| Haiku 4.5 | $0.00008 | $0.00204 |
Grade B, and why
eresus-deser-audit scanned grade B with 3 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 10d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Encoded or obfuscated payloadmediumSupply chain
base64 or hex that is decoded and executed hides what actually runs from anyone reading the file.
| `marshal.loads(data)` | High — code object creation | `json.loads()` | Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.
Unrestricted tool accesslowExcessive agency
A wildcard tool grant or "run any command" leaves no least-privilege boundary at all.
- `pickle` can execute arbitrary code with just `__reduce__` returning `(os.system, ('cmd',))` Downgraded: this mod is about security review, or the phrase is quoted, so it is likely naming the pattern rather than instructing it.
Runs shell commandslowCapability
Expected in a hook, worth knowing in a rule or an instructions file.
- The `os.system`, `subprocess.Popen` classes are directly invokable via `__reduce__` How it starts
The opening of the file, as written. The whole thing — 194 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Deserialization Vulnerability Audit
Purpose
Perform a targeted audit of deserialization attack surfaces across any language. This skill provides structured knowledge of dangerous deserialization sinks, safe alternatives, gadget chain indicators, and a step-by-step exploitation methodology.
This is a depth-first specialist skill — it goes deeper on deserialization than the
general eresus-manual-security-audit skill. Use it when the target application processes
serialized data from untrusted sources.
Universal Attack Methodology
Step 1: Identify the Deserialization Sink
Search the codebase for functions that convert serialized data back into objects.
Use grep_search with these patterns per language:
Java: ObjectInputStream, readObject, XStream, fromXML, Kryo, readClassAndObject
Python: pickle.load, pickle.loads, yaml.load, marshal.loads, shelve.open
Ruby: Marshal.load, YAML.load, Oj.load, Ox.load
PHP: unserialize, simplexml_load_string
.NET: BinaryFormatter, SoapFormatter, NetDataContractSerializer, LosFormatter
Node.js: node-serialize, funcster, cryo
Step 2: Trace the Input Source
For each sink found, trace backwards to determine:
- Does the serialized data come from an untrusted source? (HTTP request, file upload, message queue, database)
- Is there any validation or type filtering before deserialization?
- Can the attacker control the full serialized payload or only parts of it?
Step 3: Check for Type Control
The key question: can the attacker control which class/type gets instantiated?
- If the format allows arbitrary type specification (YAML tags, Java serialization, .NET TypeNameHandling), it is almost certainly exploitable
- If the format is type-restricted (JSON without polymorphism,
yaml.safe_load), it may be safe
Step 4: Identify Available Gadgets
Look for classes on the classpath/load path that have dangerous side effects during deserialization:
- Classes with
__reduce__/readObject/marshal_loadmethods - Classes that perform I/O, execute commands, or make network requests during construction
- Classes that invoke callbacks or proxy methods during deserialization
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 10d ago First seen · 194 lines · 84 tokens per session scan B 882f40303ec9
eresus-deser-audit is a skill published in the GitHub repository EresusSecurity/appsec-skills (7 stars, last pushed 5mo ago), licensed Apache-2.0. It adds 84 tokens to every session and 2,038 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it B with 3 findings (encoded or obfuscated payload, unrestricted tool access, runs shell commands). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
ai-skill-auditor
A universal, dependency-free Node.js tool to statically audit AI Skills and local agents for malicious patterns.
smart-contract-audit
Comprehensive smart contract security audit framework with multi-expert analysis. Use for full audits of Ethereum / EVM Solidity and Vyper, Solana / SVM Anchor Rust, TON / FunC / Tact, or Sui / Move projects.
infrastructure-audit
Comprehensive infrastructure security audit framework for IaC, Docker, Kubernetes, and cloud configurations. Use for full infrastructure audits.
broken-access-control
Replay captured requests with swapped identities and bumped object IDs to surface broken access control. Use when testing for IDOR or authz flaws.
jwt-attacks
Forge and re-sign captured JWTs to test signature validation - algorithm confusion (alg:none, RS256 to HS256), key injection, and secret cracking. Use when testing JWT or bearer-token auth.
ssrf-oob
Prove blind SSRF by injecting an out-of-band callback into URL params and forwarding headers, then watching for the hit. Use when testing for SSRF.