Dispatch long-running research tasks to the Manus AI agent platform from Claude Code via slash command. SessionStart + Stop hooks surface newly-completed tasks both on next session and mid-conversation (zero extra Claude turns). Retrieves file deliverables from their signed CDN links, so a task whose payload is an attachment no longer looks complete with the payload missing. On-demand polling, optional Obsidian filing, explicit-invocation cost gating.
Session handoff — capture state so any AI agent can continue your work. Auto-detects git state, plans, and environment context. Commands: /handoff:create, /handoff:quick, /handoff:resume.
OneLiteFeather's developer framework: bundles standardized tools for how the team builds software, including a knowledge graph in Outline's "Vault" collection, git hygiene for commit and PR messages, the Diátaxis documentation standard, and problem framing plus decision sparring.
Recommends the cheapest model, effort level, and execution shape that will complete a task reliably — routing atomic tasks with choose-model and decomposing multi-model work into routed execution plans with plan-execution — from versioned per-provider routing policies (Anthropic and OpenAI) and a project-local calibration log.
Cross-platform ad management for Google Ads, Meta Ads, TikTok Ads, and LinkedIn Ads. 91 tools for keyword research, campaign creation, performance analysis, and budget optimization.
Intercom integration for Claude Code. Search conversations, analyze customer support patterns, look up contacts and companies, and install the Intercom Messenger. Connect your Intercom workspace to get real-time insights from customer data.
Capture session learnings into the right Claude Code memory tier instead of dumping them all into CLAUDE.md. /notation:notate reads the session, right-sizes the flow, and presents proposals as a scorecard plus per-tier tables with diffs; /notation:notate-all applies them without the picker; /notation:feedback sends a bug or routing correction to the maintainer. The bundled notation-audit skill keeps the global CLAUDE.md under its size budget, relocating detail into linked topical notes rather than deleting it. Project-scoped and global-scoped memory are kept independent: a learning that stops being true in another repo goes to that project's memory or docs, never into your global notes, and the audit reports the two scopes separately. Backups are sorted by kind rather than lumped together: a run's own rollback copies are offered for cleanup once its verification passed, while your own CLAUDE.md snapshot history is reported and left alone - and a rotation policy of your own always wins. Capture measures rather than guesses: every proposal is priced in characters before anything is written, a whole-run budget check catches additions that are each individually fine but collectively too large, and after the writes land it re-measures and reports any edit that did not land as drafted - telling that apart from a file another session changed underneath it. Notation is additive: it preserves existing lines and dates new entries rather than overwriting history.
Marc Andreessen-mode decision and productivity skill. Market-first operator that pressure-tests ventures/ideas/features/bets (market > team > product; product/market fit is the only milestone; bias to build) and runs the 3x5-card + Anti-Todo routine. Fixed anti-sycophancy operating prompt: counterargument first, no disclaimers, explicit confidence levels, no capitulation. Issues hard verdicts backed by deterministic stdlib tools.
Marc Andreessen-mode decision and productivity skill. Market-first operator that pressure-tests ventures/ideas/features/bets (market > team > product; product/market fit is the only milestone; bias to build) and runs the 3x5-card + Anti-Todo routine. Fixed anti-sycophancy operating prompt: counterargument first, no disclaimers, explicit confidence levels, no capitulation. Issues hard verdicts backed by deterministic stdlib tools.
Marc Andreessen-mode decision and productivity skill. Market-first operator that pressure-tests ventures/ideas/features/bets (market > team > product; product/market fit is the only milestone; bias to build) and runs the 3x5-card + Anti-Todo routine. Fixed anti-sycophancy operating prompt: counterargument first, no disclaimers, explicit confidence levels, no capitulation. Issues hard verdicts backed by deterministic stdlib tools.
Marc Andreessen-mode decision and productivity skill. Market-first operator that pressure-tests ventures/ideas/features/bets (market > team > product; product/market fit is the only milestone; bias to build) and runs the 3x5-card + Anti-Todo routine. Fixed anti-sycophancy operating prompt: counterargument first, no disclaimers, explicit confidence levels, no capitulation. Issues hard verdicts backed by deterministic stdlib tools.
Marc Andreessen-mode decision and productivity skill. Market-first operator that pressure-tests ventures/ideas/features/bets (market > team > product; product/market fit is the only milestone; bias to build) and runs the 3x5-card + Anti-Todo routine. Fixed anti-sycophancy operating prompt: counterargument first, no disclaimers, explicit confidence levels, no capitulation. Issues hard verdicts backed by deterministic stdlib tools.
Scaffold shared agent context files and cross-agent memory (agent-mem) so multiple coding agents share project context and can resume each other's sessions.
Weekly Claude Code usage reports correlating ccusage spend with git commits, Claude session history, COCOMO development value estimates, and hours saved. Supports collecting data from multiple workstations (AVD, Linux jump host, local Mac) and merging into a single portable bundle for report generation.