upload-to-stitch
What this skill does
Upload local assets (images, HTML, markdown) to a Stitch project via HTTP, bypassing the MCP tool's base64 output token limits.
github/google-labs-code - Agent Snooping - 8k stars
Threat analysis
Skill info
pkg:github/google-labs-code/stitch-skills@535b088?skill=upload-to-stitchAssessments (2)
Agent Snooping
Agent Snooping via local-llm-review
SKILL.md
The skill's documentation explicitly instructs users to locate and extract API keys from user-specific configuration files (e.g., `~/.gemini/settings.json`). This is a genuine risk because it encouragOutput Handling
Output Handling via local-llm-review
scripts/upload_to_stitch.py
The script mentions that the AI model's output token limit (~16K tokens) is smaller than the base64 encoding of typical files, which could lead to truncated payloads. While this is a technical limitatBadge
Add the Anomity scan badge for upload-to-stitch to your README.
How Anomity governs this at runtime
Scan-time vetting tells you what a skill says it will do. Anomity's Endpoint Sensor sees what agents actually do: it discovers skills alongside every other AI artifact on the endpoint, and runtime governance can allow, deny, or log the tool calls a skill triggers. Policy violations route to your SIEM, Slack, email, or Jira, backed by a queryable 90-day audit trail.
Book a 30-minute demo to see your own skill inventory.
Methodology and disputes
Every skill is assessed by the Anomity Skill Intelligence engine against its public source; findings indicate risk patterns, not confirmed exploitation. Maintainer of upload-to-stitch? Report an issue or request a rescan.
