Adcp Advertising 1.0.1
v1.0.0Automate advertising campaigns with AI. Create ads, buy media, manage ad budgets, discover ad inventory, run display ads, video ads, CTV campaigns, and optim...
⭐ 0· 106·2 current·2 all-time
by@dujch
MIT-0
Download zip
LicenseMIT-0 · Free to use, modify, and redistribute. No attribution required.
Security Scan
OpenClaw
Benign
medium confidencePurpose & Capability
The name/description (advertising, campaign creation, media buying, creative management) match the SKILL.md and example code. The skill's docs and examples consistently show AdCP API calls, creative uploads, product discovery, and create/update campaign flows — all expected for an advertising automation skill.
Instruction Scope
The SKILL.md and companion docs instruct the agent to call external AdCP endpoints (test and production agent URLs), use an AdCP client, upload creatives by URL, and create/modify campaigns. The instructions do not ask the agent to read local system files, unrelated environment variables, or private credentials beyond auth tokens for AdCP. Note: examples import '@adcp/client/testing' and reference an 'agent' object — the skill assumes client libraries or a runtime tool surface are available, which is an implementation assumption rather than an explicit dependency.
Install Mechanism
No install spec or code to download — it is instruction-only. No archives, URLs, or binary installs are present in the package metadata.
Credentials
The skill declares no required env vars or credentials. The documentation includes a public test-agent URL and a shared test bearer token (explicitly labeled 'intentionally public') for development/testing. That is plausible for an API-centric skill, but you should be aware the skill encourages using that public token and will make network calls. Production operations do require bearer tokens per the docs; those would be sensitive and should be stored securely by the user.
Persistence & Privilege
The skill does not request 'always: true' and makes no claims about modifying other skills or system-wide settings. Autonomous model invocation is allowed (platform default) but not excessive by itself.
Scan Findings in Context
[no-findings] expected: The static scanner had no code to analyze because this is an instruction-only skill composed of documentation and examples. Absence of findings is expected but not evidence of safety — behavior is governed by runtime instructions that call external endpoints.
Assessment
This skill is documentation-heavy and appears to implement an advertising agent that calls remote AdCP endpoints. Before installing: (1) be prepared for the agent to make outbound network requests to the test and production AdCP endpoints (the README includes a public test token and test URLs); (2) do not store production bearer tokens in plaintext in your workspace — use secure env/secret managers; (3) if you want to avoid any external calls, do not enable autonomous invocation or disable the skill until you review how your agent runtime implements the 'agent' client calls; (4) verify you trust the AdCP endpoints (docs.adcontextprotocol.org / test-agent.adcontextprotocol.org) and the repository source before providing any production credentials; (5) if you have an internal security policy, have the skill reviewed by your network/security team and monitor the agent's outbound traffic during initial use.Like a lobster shell, security has layers — review code before you run it.
latestvk97fm2f3n7b92td1c72p9245an838p40
License
MIT-0
Free to use, modify, and redistribute. No attribution required.