# Install-time and editor-config supply chain execution

Commits or packages that add auto-executing config, hooks, or obfuscated modules reaching remote payloads

_Area: Web infrastructure and server-side requests. Seen in 3 disclosed reports; latest 2026-07-01. Updated 2026-10-24._

## How it works
Code that runs automatically when a developer opens a folder, installs a package, or applies a setting is treated as inert data. The assumption that config files, editor directories, and third-party modules are passive breaks: they can carry executable logic that fetches and runs remote payloads. The trust boundary is the developer's machine, and it is crossed without any explicit action beyond opening or installing.

## Where to look
- Commits adding `.claude/`, `.gemini/`, `.cursor/rules/`, `.vscode/tasks.json` or similar agent/editor config that auto-executes on open [E1](https://bugbountylibrary.com/r/blog%3Astepsecurity-io-blog-miasma-worm-hits-microsoft-again-azure-functions-action-and-72-other-repositories-disabled-after-su.md)
- Commits whose message claims a code change but touches zero source files, is backdated, or carries `[skip ci]` [E1](https://bugbountylibrary.com/r/blog%3Astepsecurity-io-blog-miasma-worm-hits-microsoft-again-azure-functions-action-and-72-other-repositories-disabled-after-su.md)
- Third-party modules importing `codecs`+`base64`+`subprocess`+`tempfile` together, or holding large Base64 string constants [E2](https://bugbountylibrary.com/r/blog%3Aelastic-co-security-labs-threat-command-dprk-code-of-conduct.md)
- ROT13-looking hostnames (e.g. `uggcf`) and `exec()` of server-returned data [E2](https://bugbountylibrary.com/r/blog%3Aelastic-co-security-labs-threat-command-dprk-code-of-conduct.md)
- Plugin `save()`/apply handlers that read API keys and POST them to hardcoded IPs [E3](https://bugbountylibrary.com/r/blog%3Astepsecurity-io-blog-jetbrains-malicious-plugins-ai-api-key-theft.md)
- JVM-wide `X509TrustManager` overrides suppressing TLS warnings [E3](https://bugbountylibrary.com/r/blog%3Astepsecurity-io-blog-jetbrains-malicious-plugins-ai-api-key-theft.md)
- READMEs urging users to run something before editing [E2](https://bugbountylibrary.com/r/blog%3Aelastic-co-security-labs-threat-command-dprk-code-of-conduct.md)

## How to review
For every added config directory, hook, or dependency, ask: does this execute on open/install, and where does its payload come from? The invariant: no auto-run path may reach a network fetch or `exec` without explicit user consent and a pinned, verified source. Trace each outbound request to a named, expected host; flag raw IPs and obfuscated strings. Confirm TLS validation is not globally disabled. Check that the commit's stated intent matches its actual file changes.

## Variants seen
- Editor/AI-agent config auto-executing a credential harvester on folder open [E1](https://bugbountylibrary.com/r/blog%3Astepsecurity-io-blog-miasma-worm-hits-microsoft-again-azure-functions-action-and-72-other-repositories-disabled-after-su.md)
- Trojanized Python module decoding and exec'ing remote payloads [E2](https://bugbountylibrary.com/r/blog%3Aelastic-co-security-labs-threat-command-dprk-code-of-conduct.md)
- Marketplace plugin intercepting API keys on Apply and exfiltrating over plaintext HTTP [E3](https://bugbountylibrary.com/r/blog%3Astepsecurity-io-blog-jetbrains-malicious-plugins-ai-api-key-theft.md)

## Impact and severity
High: credential and API-key theft at scale (70,000 installs [E3](https://bugbountylibrary.com/r/blog%3Astepsecurity-io-blog-jetbrains-malicious-plugins-ai-api-key-theft.md)), worm-like propagation across 72+ repositories [E1](https://bugbountylibrary.com/r/blog%3Astepsecurity-io-blog-miasma-worm-hits-microsoft-again-azure-functions-action-and-72-other-repositories-disabled-after-su.md), and RCE on developer workstations [E2](https://bugbountylibrary.com/r/blog%3Aelastic-co-security-labs-threat-command-dprk-code-of-conduct.md). Severity rises with auto-execution on open, plaintext exfil, and reuse of compromised tokens that let the payload spread.

## Evidence
- E1: https://bugbountylibrary.com/r/blog%3Astepsecurity-io-blog-miasma-worm-hits-microsoft-again-azure-functions-action-and-72-other-repositories-disabled-after-su.md
- E2: https://bugbountylibrary.com/r/blog%3Aelastic-co-security-labs-threat-command-dprk-code-of-conduct.md
- E3: https://bugbountylibrary.com/r/blog%3Astepsecurity-io-blog-jetbrains-malicious-plugins-ai-api-key-theft.md