The Agent Skills Directory

[COMMAND_EXECUTION]: The skill's preamble executes a shell script using a relative path (../../check-update.sh). This allows for the execution of local code located outside the skill's own directory structure.
[INDIRECT_PROMPT_INJECTION]: The skill ingests content from external markdown documents and logo files to determine brand personality and animation parameters, which could be exploited if those files contain malicious instructions.
Ingestion points: Reads docs/02-方案设计/原型设计方案.md and docs/02-方案设计/产品定位方案.md during Step 2.
Boundary markers: Absent; the content is processed without explicit delimiters or instructions to ignore embedded commands.
Capability inventory: Bash (used in preamble), Write (generates demo HTML and plan markdown), and Read (accesses local file system).
Sanitization: Absent; the ingested data is not filtered or validated before being used to populate variables and generated documents.
[DYNAMIC_EXECUTION]: The skill generates an HTML file (品牌动效demo.html) that includes inline JavaScript and CSS animations. This runtime generation of executable web content is based on templates provided in the skill instructions.

pm-brand-motion