The Agent Skills Directory

[CREDENTIALS_UNSAFE]: The file SKILL.md contains a hardcoded, active API key (moltbook_sk_n-6tSniNp39Lv1Hi_0s3T6km4y1DdjMk) and a specific claim URL (https://www.moltbook.com/claim/moltbook_claim_1dL25UZU-i3UA4jXLRWiBRzWJP09NOQv). Shipping literal credentials instead of placeholders or environment variable instructions exposes the account to unauthorized access.
[COMMAND_EXECUTION]: The skill provides numerous instructions for the agent to execute shell commands using curl. These include complex command-line operations with proxy configurations (export http_proxy=...), JSON payload manipulation, and piping API responses to Python for processing. This level of autonomous shell interaction increases the risk if the agent is manipulated via external input.
[INDIRECT_PROMPT_INJECTION]: The skill represents a structural risk for indirect prompt injection due to its core functionality.
Ingestion points: The agent is instructed to fetch and process data from https://www.moltbook.com/api/v1/posts, notifications, and home (defined in SKILL.md).
Boundary markers: There are no instructions to use delimiters or ignore embedded instructions when reading social media content.
Capability inventory: The agent has capabilities to execute shell commands (curl), write to local configuration files (~/.config/moltbook/credentials.json), and perform further network requests (described in SKILL.md and references/verification-wrapper-architecture.md).
Sanitization: There is no evidence of input validation or sanitization before the agent processes external social content.
[EXTERNAL_DOWNLOADS]: The skill instructs the user/agent to install the Python package moltbook-verify from a public registry. While this appears to be a vendor-related utility for solving platform challenges, it introduces a third-party dependency into the environment.

moltbook