Agent Package

A repeatable daily rhythm for solopreneurs and operators using Claude Code as an always on chief of staff. Inspired by Jim Prosser (@jimprosser): "My chief of staff, Claude Code". When to use: Every working day. Adapt the cadence to your schedule. Preconditions: Claude Code is configured with access to your calendar, task list, and inbox (via MCP servers or file based context) A CONTEXT.md or equivalent daily state file is maintained in your working directory You have a GOALS.md file Claude Code references for alignment

Small Feature Ship Workflow A reusable workflow for shipping small product or engineering features with minimal human coordination and clear agent ownership. This workflow is optimized for small, bounded changes where the goal is to avoid process overhead while still protecting product judgment, implementation quality, and release safety. How to Use When a user makes a small feature/change request in chat, treat this workflow as the operating procedure and execute s001 → s004 in the same chat session. Intermediate artifacts (decision brief, adversarial review prompt, Codex verdict) live in the chat context — no external persistence needed. The final deliverable returned to the user in chat is the PR URL . Role Map | Role | Assignee | Responsibility | | | | | | Feature Owner | human | Owns intent, final judgment, and shipping decision. | | Decision Challenger | Claude Code | Challenges whether the feature should be built, whether the proposed solution is the right one, and what the real problem is. | | Builder | Claude Code | Implements the agreed solution, opens the PR. | | Reviewer | Codex (preferred) / Claude Code (fallback) | Adversarially reviews the PR. Codex owns the verdict when available; otherwise Claude Code performs the same adversarial review against its own work. | | Shipper | human | Reviews the PR, merges/releases, and observes early results. | Core Loop Decide → Build → Check → Ship Four steps. Each step has a clear owner, with explicit handoff artifacts so the next owner can proceed without ambiguity. Operating Principle Do not move forward simply because a feature request exists. The first step must pressure test the request — but right size the scrutiny to the change . A variable rename does not deserve a 9 point brief; a behavior change with user impact does. The Decide step defines what "enough scrutiny" looks like for each tier.

Purpose This workflow defines a structured approach to introducing an AI agent into business operations, clarifying its role, preparing the required data environment, and establishing continuous evaluation and improvement mechanisms. It ensures that the AI agent delivers reliable, high accuracy outputs while remaining aligned with real business needs. Scope The workflow covers: Decomposition of existing business tasks and scenarios suitable for AI handling Data preparation and governance to support accurate AI reasoning Continuous evaluation and feedback loops for quality improvement Collaboration structures between business and IT teams to enable sustainable operation Outcome By following this workflow, organizations can: Clearly define what the AI agent should and should not do Improve AI accuracy through better data and feedback cycles Reduce misalignment between business expectations and AI behavior Establish a repeatable, scalable model for AI agent deployment

A weekly system that turns one “core idea” into a high quality long form asset (newsletter or blog post), then repurposes it into short form social content (X thread, LinkedIn post, atomic posts). A searchable personal “Content Vault” (for example, Manus) makes the workflow faster by letting you retrieve your best prior stories, frameworks, and phrases on demand. References: New Manus 1.6 Is Cracked (Absorb Years Of Your Best Thinking In Minutes) Weekly goal 1 long form piece (publish ready) 10 to 20 short form assets derived from it Scheduled distribution for the week Weekly deliverables Newsletter or blog post: 1 X thread: 1 LinkedIn post: 1 Atomic posts: 5 to 10 Scheduling plan: 1 week

Agent Package is a shared system for turning successful AI work into reusable operational assets. It gives users two core building blocks: Workflow Pack (reusable execution) and Context Pack (portable understanding). Designed for personal use first, but built to support team sharing and community distribution. Core idea: useful AI work should not disappear into chat history. It should become portable, inspectable, reusable, and continuously improvable.

Most projects fail not from lack of effort, but from lack of shared understanding — of goals, assumptions, and what done actually means. Epismo is built to fix that. By structuring the layer before tasks exist and making that structure accessible to both humans and AI agents, Epismo enables human AI teams to move faster without losing alignment. This pack covers the philosophy behind Epismo, what it is built on, and what makes it different.

A pressure test for an unvalidated idea before any building begins. Six forcing questions, asked one at a time, that surface hidden assumptions, refuse polite agreement, and convert vague conviction into a small, falsifiable next step. Why this pack exists Most ideas die from unexamined assumptions, not from bad execution. When a builder pitches an idea — to a friend, a partner, an AI assistant — the default response is some version of "interesting!" That politeness is the most expensive thing in early stage product work. It lets bad ideas survive long enough to consume months of building. This pack replaces the polite mirror with a structured interrogation. The six questions are sequenced so each one closes off a common escape route: vague users, hypothetical workarounds, missing "why now," unexamined beliefs, optimism bias, and the urge to start coding instead of testing. Core idea Specificity is the evidence of understanding. A builder who can name one real person, one real moment, and one real workaround has done the work. A builder who speaks in plurals and abstractions has not. Every question in this pack pushes relentlessly toward the singular and concrete. Who this is for Anyone holding an idea that has not yet met reality — solo builders, side project starters, founders early in formation, internal product owners pitching a new initiative. Not for products that already have paying customers. Not for technical implementation questions. Not for emotional support sessions. When to use it Load this pack when the user signals an unvalidated idea: "I'm thinking about building...", "is this worth doing?", "what do you think of this?", or any pitch of a half formed concept. Do not load when there are already paying customers, when the question is about implementation, when the user is venting, when the decision is irreversible, or when this same idea has been stress tested in a recent session. Expected duration 20–40 minutes of back and forth. Faster usually means worse. The output is a written diagnostic with a single verdict: GO, VERIFY, RECONSIDER, or KILL — plus a 7 day falsifiable test.

A fat, AI ready playbook for turning messy customer evidence into decision quality insight. Based on the pm skills market research methodology: segmentation, personas, customer journey mapping, competitor analysis, market sizing, sentiment analysis, and feedback synthesis. Each block is a discrete skill — use them together for a full research cycle or independently as targeted methods. The goal is always to produce research that changes a decision, not research that fills a slide deck.