# Hunt Deserialization

> Investigate authorized unsafe deserialization and serialized-data trust boundaries. Use when typed discovery identifies encoded object cookies or tokens, pickle or object streams, YAML type tags, serialized uploads or imports, type-confusion errors, or suspected server-side object reconstruction.

- Skill: `duriantaco/hunt-deserialization` (Agent Skill, multi-file: 2 files)
- Install (CLI): `npx skillmds@latest add duriantaco/hunt-deserialization`
- Raw SKILL.md: https://api.skillmd.com/api/skills/duriantaco/hunt-deserialization/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: Product & Planning
- Author: duriantaco (https://skillmd.com/u/duriantaco)
- Updated: 2026-09-17
- Page: https://skillmd.com/skills/duriantaco/hunt-deserialization

---


# Hunt Deserialization Boundaries

Treat this card as advisory prioritization. Encodings, format fingerprints, and parser errors do not
prove unsafe object construction.

## Workflow

1. Identify the format, carrier, integrity protection, identity, and target-observed replay.
2. Pair an unchanged baseline with a same-shape inert structural control; change one property.
3. Use `cookie_deserialization` for cookies or tokens and `file_fetch_parser` only for typed imports
   or uploads. Their fixed command and file-read candidates must all be authorized.
4. Preserve redacted inputs, format decisions, controls, deltas, and evidence references. Report a
   validated vulnerability independently of any challenge objective.

## Evidence Gate

Confirm only a replayable unsafe type, object-handling, or execution effect absent from the paired
control. Decoder errors, library names, gadget strings, and status changes remain observations.
Keep any `contract_missing` result suspected rather than promoting it.

## Stop Conditions

Stop when format, integrity, identity, replay, or controls are unavailable, or policy limits fire.
Do not add destructive gadgets, persistence, or outbound effects; redact returned secrets.