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 MingyiSecLab/Mingyi-Atlas --skill prototype-pollutiongit clone --depth 1 https://github.com/MingyiSecLab/Mingyi-AtlasWrote 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/mingyiseclab/mingyi-atlas/prototype-pollution)<a href="https://agentmods.dev/skills/mingyiseclab/mingyi-atlas/prototype-pollution"><img src="https://agentmods.dev/badge/skills/mingyiseclab/mingyi-atlas/prototype-pollution/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/mingyiseclab/mingyi-atlas/prototype-pollution"><img src="https://agentmods.dev/badge/skills/mingyiseclab/mingyi-atlas/prototype-pollution.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 | $0.00047 | $0.01159 |
| Opus 5 | $0.00023 | $0.00580 |
| Sonnet 5 | $0.00009 | $0.00232 |
| Haiku 4.5 | $0.00005 | $0.00116 |
Grade A, and why
prototype-pollution 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 9d 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.
curl -X POST https://target.com/api/settings \ Runs shell commandslowCapability
Expected in a hook, worth knowing in a rule or an instructions file.
- **`child_process.spawn(cmd, args, opts)`** — opts has a `shell` option. This is a copy
88% identical to prototype-pollution — 3 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
How it starts
The opening of the file, as written. The whole thing — 104 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Prototype Pollution Playbook
Prototype pollution is the JavaScript equivalent of a universal gadget:
plant a property on Object.prototype and it appears on every object in
the runtime. Worthless in isolation, deadly in chain (__proto__.isAdmin = true → auth bypass; __proto__.shell = "/bin/bash" → RCE via spawn).
1. Sinks — the libraries that still introduce sinks
Keep a running list per engagement. These continue to ship sinks in 2026:
- Deep-merge:
lodash.merge,deepmerge(pre-fix),merge-deep,deepExtend,hoek.merge,mixme - Deep-clone:
lodash.defaultsDeep,lodash.zipObjectDeep,set-value(pre-3.0.3) - URL-to-obj:
qs,express-fileupload,jquery.extend(true, ...) - Config loaders:
node-configrecursive merge,dotenv-extended,rc - Template engines: Handlebars helpers fed from untrusted ctx
# Every JS/TS project: sweep known-bad versions
jq '.dependencies,.devDependencies | to_entries[] | select(.key | test("merge|lodash|set-value|dot-object|dot-prop|node-pg"))' /workspace/src/package.json
npm ls lodash set-value dot-prop 2>/dev/null | grep -E '[0-9]'
2. Sources
Any user input deserialized into a nested object:
- JSON body parsers (
body-parser,express.json) - Query string parsers (
qswith default config parsesa[b][__proto__][c]=1) - YAML uploads
- Form-data / multipart
3. Audit workflow
- Find every deep merge call site.
- Trace each one backwards — is the right-hand object user-controlled?
- If yes: check the merge function's prototype-pollution fix version.
- Even if the merge is fixed, check whether a copy (lodash.set, dot-path-value, jsonpath.set) creates a pollution path.
4. Exploitation gadgets
Poisoning Object.prototype doesn't do anything by itself — you need
a gadget that reads a property that didn't exist before.
Classic gadgets:
child_process.spawn(cmd, args, opts)— opts has ashelloption. Poison__proto__.shell = "/bin/bash"then any subsequent spawn call executes through bash and interprets args as shell strings.- Express middleware — most middlewares check
options.someFlagwithif (opts.someFlag). Poisoning that flag flips security defaults. - Templating — Handlebars and EJS read
helpersandpartialsfrom the context object; pollution adds helpers that execute code. lodash.template— if the template source is built from_.template(tpl, ctx)you can inject via pollutedescape/evaluatekeys.mongoose— pollutingSchema.Typescauses subsequent schema definitions to use attacker-controlled types.
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.
- 9d ago First seen · 104 lines · 47 tokens per session scan A b0672a54794d
prototype-pollution is a skill published in the GitHub repository MingyiSecLab/Mingyi-Atlas (11 stars, last pushed 2mo ago), licensed Apache-2.0. It adds 47 tokens to every session and 1,159 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). It is 88% identical to prototype-pollution, differing in 3 lines, and is treated as a copy.
Other skills, from other repositories
cis-aws-foundations-4.3
Ensure AWS Config is enabled in all regions.
cis-aws-foundations-6.5
Ensure the default security group of every VPC restricts all traffic.
cis-aws-foundations-2.1.3
Ensure Organizations management account is not used for workloads.
cis-aws-foundations-2.5
Ensure MFA is enabled for the 'root' user account.
cis-aws-foundations-2.7
Eliminate use of the 'root' user for administrative and daily tasks.
cis-aws-foundations-4.4
Ensure that server access logging is enabled on the CloudTrail S3 bucket.