fellowship-dev
- 56 skills
- 0 followers
- 11 hours ago last updated
- ▌ Navvi Login · fellowship-devLog into a web service using Navvi — reads gopass credentials, fills login form, handles reCAPTCHA and 2FA escalation. Use when asked to log in, sign in, or authenticate with a website.
- ▌ Navvi Browse · fellowship-devAutonomous browser agent — navigates, interacts, and reports using Navvi MCP tools. Use when asked to browse a website, interact with web pages, scrape content, or perform any web browsing task.
- ▌ Navvi Signup · fellowship-dev bundleCreate a new account on a web service using Navvi — generates credentials in gopass, fills signup form, handles CAPTCHAs, registers the account. Use when asked to sign up, create an account, or register on a website.
- ▌ Flowchad · fellowship-dev bundleDrop-in AI QA for any web project — automated flow walking, screenshot capture, and regression detection
- ▌ Flow Add · fellowship-devCreate a new flow definition from a natural language description — scans codebase for routes, components, and existing flows to generate accurate YAML. Usage /flow-add <description>
- ▌ Flow Diff · fellowship-devCompare walk snapshots to detect regressions and improvements — visual diff, timing changes, broken/fixed steps. Usage /flow-diff <flow-name>
- ▌ Flow Walk · fellowship-devWalk a user flow using Playwright CDP — execute steps, capture screenshots, measure timing, store results. Usage /flow-walk <flow-name>
- ▌ Flow Report · fellowship-devGenerate a categorized friction report from walk results — Critical/Friction/Cosmetic findings with actionable suggestions. Usage /flow-report <flow-name>
- ▌ Flow Update · fellowship-devUpdate an existing flow definition to reflect product changes — reads codebase diffs, updates steps/expectations/context. Usage /flow-update <flow-name> <what changed>
- ▌ Flow Diagram · fellowship-devGenerate mermaid flowcharts from YAML flow definitions — visualize happy path, error paths, and decision points. Usage /flow-diagram <flow-name>
- ▌ Flow Suggest · fellowship-devAI-powered improvement prioritization — analyze walk results, friction reports, and trends to produce a ranked action plan. Usage /flow-suggest <flow-name>
- ▌ Flow Reproduce · fellowship-devReproduce a customer-reported bug in-browser given a GitHub issue with error details. Walks an inferred flow, captures evidence, and updates the original issue with reproduction outcome + steps-to-reproduce. Usage /flow-reproduce <issue-ref>
- ▌ Flowchad Setup · fellowship-devInitialize Flowchad in a project — auto-discover existing tests, analytics, specs, and user flows, then scaffold flow definitions. Run as first step before any /flow-walk.
- ▌ Fly Io · fellowship-dev bundleUse when deploying to Fly.io or configuring flyctl, fly.toml, Fly.io secrets, databases, or GitHub Actions.
- ▌ SEO Ops · fellowship-dev bundleUse when auditing a website before launch or in production for canonical hosts, redirects, crawlability, indexability, multilingual hreflang, metadata, structured data, sitemaps, robots, internal links, and recurring SEO regressions.
- ▌ Hookshot · fellowship-devUse when generating Claude Code enforcement hooks from a repo's docs/ structure.
- ▌ Popsicle · fellowship-dev bundleUse when generating agent-native onboarding docs, coverage maps, health baselines, and agent adapters for a repo.
- ▌ Pylot CLI · fellowship-dev bundleUse when operating the Pylot gateway through its CLI — dispatch, monitor, secrets, assets, workers, and automations.
- ▌ Review Pr · fellowship-dev bundleUse when performing a first-pass, read-only PR review in a clean isolated context.
- ▌ Cto Review · fellowship-dev bundleUse when performing a CTO-level PR review — includes a staging evidence gate for release-train PRs (base = default branch; per-PR staging retired 2026-09-06, pylot#3389) and a non-blocking visual evidence notice for UI PRs.
- ▌ Dx Metrics · fellowship-dev bundleUse when measuring AI-coding-agent adoption and its effect on software delivery for a repo or org — leverage per human steerer, DX Core 4 speed/quality/impact, AI-execution floors, cycle time, ownership risk — computed from git and the GitHub API into a self-contained HTML dashboard. Attributes every commit to a steerer AND an executor rather than classifying it human-or-AI.
- ▌ Playwright · fellowship-devUse when automating a browser with Playwright CLI over CDP — navigate, screenshot, fill forms, or extract data.
- ▌ Vercel Ops · fellowship-devUse when safely inspecting or changing an existing Vercel project's environment variables, root-directory link, preview deployments, deployment protection, or custom domains.
- ▌ Build Train · fellowship-dev bundleUse when batching multiple GitHub issues into one build branch with concurrent subagent builds.
- ▌ Deps Runner · fellowship-dev bundleUse when running the full dependency update pipeline for a repo — scan, risk-eval, build-test, and merge-decision.
- ▌ Docs Review · fellowship-devUse when checking for drift between docs/ and source code — reports discrepancies with file:line evidence.
- ▌ Maintenance · fellowship-devUse when running an infra-only health audit across Pylot systems — LaunchAgent, cron, secrets, labels, branches, and issues.
- ▌ Trash Truck · fellowship-dev bundleInvestigate and retire unused product, code, operational, or infrastructure surfaces using real usage evidence. Use when a repository needs active pruning, when a specific capability may no longer be necessary, or when a scheduled review should preserve retirement candidates without executing them. Do not use for routine lint cleanup, broad refactoring, or dependency updates.
- ▌ Weekly Plan · fellowship-devUse when running the interactive weekly planning session for a Pylot team (interactive session only — not headless).
- ▌ Daily Report · fellowship-devUse when writing the daily/rollcall team report — defines 5 sections, quality rules, and a minimal idle card.
- ▌ Double Check · fellowship-dev bundleUse when performing a standalone PR double-check in a clean context — review, fix, and post.
- ▌ Issue To Prd · fellowship-dev bundleUse when converting a new GitHub issue into a structured PRD, or posting clarifying questions when the issue is underspecified.
- ▌ Setup Github · fellowship-dev bundleUse when setting up a GitHub project board, labels, and Actions workflows for a new or existing repo.
- ▌ Write Report · fellowship-devUse when writing a mission report to the gateway DB from any operator mission.
- ▌ Entropy Check · fellowship-devUse when checking doc freshness and computing domain quality grades without making fixes.
- ▌ Pylot Workers · fellowship-devDEPRECATED — worker spawn/drive/stop moved into pylot-cli. Read pylot-cli instead; this skill is a pointer and carries no instructions.
- ▌ Setup Harness · fellowship-devUse when scaffolding the agent knowledge layer (ARCHITECTURE.md, QUALITY_SCORE.md, docs/) for a repo.
- ▌ Setup Speckit · fellowship-devUse when setting up speckit (Specification-Driven Development) in a repo — copies slash commands, templates, and optionally creates a project constitution.
- ▌ Vercel Deploy · fellowship-devUse when deploying an existing Vercel project to production non-interactively.
- ▌ Build Prd · fellowship-devUse when creating a PRD, writing requirements, or transforming a feature request into an implementation-ready GitHub issue.
- ▌ Spec Plan · fellowship-devUse when stress-testing a plan, getting grilled on a design, or before writing a PRD.
- ▌ Migrate Skill · fellowship-devUse when consolidating a skill from a claude-toolkit plugin or local .claude/skills into the dogfooded-skills library.
- ▌ Skill Builder · fellowship-devUse when writing a new agent skill for the dogfooded-skills library — covers frontmatter spec, section structure, quality criteria, and antipatterns.
- ▌ Security Check · fellowship-devUse when triaging Dependabot/Snyk security alerts by severity and exploitability before running security-runner.
- ▌ Speckit Runner · fellowship-dev bundleUse when implementing a GitHub issue end-to-end — runs speckit with an independent advisory review before creating one PR.
- ▌ Agent Readiness · fellowship-devUse when assessing whether a repo is ready for autonomous AI agents (pylot workers) — produces a scored, actionable Agent Readiness Report as a GitHub issue in the assessed repo.
- ▌ Evidence Upload · fellowship-devUse when publishing visual evidence through Pylot assets for a PR, issue, report, mission, or conversation.
- ▌ Flowchad Runner · fellowship-dev bundleUse when verifying local, preview, production, or recurring critical user flows with browser evidence and selective on-demand deployments.
- ▌ Security Runner · fellowship-devUse when running the automated security alert fix pipeline — opens fix PRs, creates issues for breaking changes, dismisses false positives.
- ▌ Visual Evidence · fellowship-devUse when capturing Playwright screenshots or GIF recordings for PR evidence — before/after comparisons and feature demos.
- ▌ Refactor Codebase · fellowship-dev bundleAutonomously refactor a codebase across many verified cycles — decompose fat files by domain, deepen shallow modules, pay down architectural debt — driven by a persistent masterplan and behavior-preserving gates. Use when the user wants to refactor at scale, split god-files, decompose a large module, run an ongoing architecture-improvement loop, or do a multi-session refactor that must not change behavior. Complements (does not replace) improve-codebase-architecture, which only finds and reports opportunities.
- ▌ Procedure Builder · fellowship-dev bundleUse when building a new multi-stage ICM procedure or converting a monolithic runner skill into a staged workflow.
- ▌ Setup Devcontainer · fellowship-devUse when generating devcontainer config or creating a Gitpod/Ona project for any repo.
- ▌ Requestor Attribution · fellowship-devRead at session start — injects Co-authored-by trailers and a "Requested by" PR attribution line from $PYLOT_REQUESTORS. No-op when the env var is absent or empty.
- ▌ Create Compelling Prs · fellowship-devUse when preparing a PR for review — applies body templates, attaches the deployment and visual evidence the repo playbook requires, and runs the self-audit checklist.
- ▌ Create Compelling Issues · fellowship-devUse when filing a GitHub issue — search for duplicates first, resolve filing policy from the repo playbook, and ground every claim in evidence.