The systems I use to run SEAR Plugs alongside a full-time career — documented one by one. Each entry follows the same structure: the problem, the old manual process, the AI-assisted version, where human review is required, the outcome, and what failed. No invented numbers; where results aren't measured yet, it says so.
Content Systems
LiveFive purpose-built Claude skills and two scheduled tasks run SEAR's blog — planning separated from writing, with human gates at every publish decision.
Tools: Claude (Cowork) skills, Google Search Console MCP, WordPress MCP, BigQuery, Scheduled tasks
View on GitHubWebsite Optimisation
LiveEvery Monday at 07:00, a scheduled audit pulls Search Console signals, checks them against declarative rules, and emails a prioritised action list — with every finding logged and deduplicated in BigQuery.
Tools: Claude (Cowork) skills, Google Search Console MCP, WordPress MCP, BigQuery, Scheduled tasks
View on GitHubPaid Social
LiveA sequenced monthly Meta Ads review — fatigue, audience overlap, budget allocation — run through a Meta Ads MCP and kept deliberately bounded, so it ends in decisions rather than dashboards.
Tools: Claude (Cowork), Meta Ads MCP, Meta Marketing API
View on GitHubConversion
LiveA repeatable process for auditing and rewriting product pages — meta, structure, copy — while keeping SEAR's voice intact.
Tools: Claude (Cowork), WooCommerce, Google Search Console
View on GitHubA note on scope: everything documented here comes from SEAR or personal experiments. Workflows from my employer's environment stay there — I write about enterprise AI adoption only at the level of general, publicly discussable patterns.