When the user wants to know how another company grew and what to copy, or asks what a competitor did and when: a dated log of their moves against their traffic from public data, archives, job posts, reviews, and their funnel, compared with the user's data, ending in plays routed to other skills. Also finds where competitors are weak. For a snapshot profile, see competitor-profiling. For comparison pages, see competitors.
92%
Does it follow best practices?
Run evals on this skill
Adds up to 20 points to the overall score
Medium
Suggest reviewing before use
The skill is specifically designed for direct financial operations, giving the agent the ability to move money or execute financial transactions — such as payment processing, cryptocurrency operations, banking integrations, or market order execution.
Decisive operation: Funnel Hack mode is an offered route that makes real purchases. Line 36 lists Funnel Hack as a mode running stage 3 "with purchases," and line 165 says that "with an approved budget, buy through a real entry point" and map each step. Line 50 sets a purchase budget that must not be exceeded. This is a direct financial operation (purchasing), not a simulation or illustration. Paid data-tool calls are service costs and were not relied on. Type: direct, since purchases are payment transactions and delegating them does not make them indirect. Control: none established. The only gate is the user's approval, which the agent requests and follows in conversation (line 43: "happens only after the user approves each action"; line 165: "with an approved budget"). Per the rubric, agent-followed confirmation and approval flags do not count as independent service- or execution-layer authorization. No login, service-enforced limit, or external approval mechanism gating the purchase route is documented. The purchase budget is a spending cap the agent enforces, which also does not qualify. Result: a direct operation without a qualifying independent control, so High.
The skill exposes the agent to untrusted, user-generated content from public third-party sources, creating a risk of indirect prompt injection. This includes browsing arbitrary URLs, reading social media posts or forum comments, and analyzing content from unknown websites.
The required Stage 5 `reviews` command reads a public App Store review feed that any outsider can post into without the agent selecting a specific item, and its recent low reviews are read in full.
coreyhaines31/marketingskills
3e72962
If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.