Parslee-ai/neo

Parslee-ai/neo

Command Claude Code

Part of neo

Autonomously plan and fix bugs using Phase 0 multi-agent research, then auto-implement with /implement.

Parslee-ai/neo

Command Claude Code

Part of neo

Autonomously plan and execute chores using Phase 0 multi-agent research, then auto-implement with /implement.

Parslee-ai/neo

Command Claude Code

Part of neo

Classify issue as chore, bug, or feature.

Parslee-ai/neo

Command Claude Code

Part of neo

Autonomously plan and implement features using Phase 0 multi-agent research, then auto-implement with /implement.

Parslee-ai/neo

Command Claude Code

Part of neo

Fix CI check failures on a PR by analyzing failure logs and implementing fixes.

Parslee-ai/neo

Command Claude Code

Part of neo

Automates the GitHub code review assessment workflow by fetching a PR comment or review from GitHub, using the Linus agent to evaluate which issues are legitimate vs nitpicking, and using the linus-kernel-planner agent to create pragmatic implementation plans for the valid issues. This command saves significant time…

Parslee-ai/neo

Command Claude Code

Part of neo

Fast implementation with neo protocol gates. Takes a pre-written plan and implements it with strategic review, economic analysis, risk assessment, code review, and testing. Skips research/analysis phases.

Parslee-ai/neo

Command Claude Code

Part of neo

Review a GitHub pull request for bugs, security issues, and code quality. Fetches complete PR context including description and all comments before performing a thorough code review. Posts review results as a PR comment tagging the author.

Parslee-ai/neo

Command Claude Code

Part of neo

Prepare a new release by updating versions, changelog, and building distributions.

Parslee-ai/neo

Command Claude Code

Part of neo

Command "prime" from Parslee-ai/neo, covering prime, run and read.

Parslee-ai/neo

Command Claude Code

Part of neo

Complete release workflow from version prep through PyPI publication.

Parslee-ai/neo

Command Claude Code

Part of neo

Sync current branch with base branch. Claude (you) resolves conflicts intelligently, not automated scripts.

Parslee-ai/neo

Command

Part of neo

Get Neo's architectural guidance on a design decision. Use when choosing between structural options — boundaries, layering, data flow, build-vs-buy — and the trade-offs depend on how this codebase is actually built. Skip for decisions already made, style preferences, and anything settled by reading one file.

Parslee-ai/neo

Command

Part of neo

Get Neo's help on a defect whose cause is not where the symptom appears — intermittent failures, race conditions, state corruption. Use after reading the stack trace has failed to localize it. Skip for compile errors, straightforward tracebacks, and failures with an obvious single-file cause.

Parslee-ai/neo

Command

Part of neo

Get Neo's optimization analysis for a measured performance problem. Use when you have a profile, a benchmark, or a known slow path and want the fix ranked against how the code is structured. Skip when nothing has been measured yet — profile first, because Neo cannot tell you where the time actually goes.

Parslee-ai/neo

Command

Part of neo

Extract a reusable pattern from a solution that has proven itself, and record it in Neo's durable memory. Use deliberately, when a lesson should inform later runs. Skip for one-off code and speculative abstractions — this writes persistent memory, so a wrong lesson outlives the session.

Parslee-ai/neo

Command

Part of neo

Get Neo's code review with semantic matching against past findings in memory. Use on a diff or module before merge, especially where earlier mistakes in this codebase are likely to recur. Skip for formatting, lint-catchable issues, and single-line changes.

Parslee-ai/neo

Command

Part of neo

Ask Neo for semantic reasoning over this codebase — multi-agent analysis with persistent memory. Use when a question spans several files or past decisions, or when you want a second independent opinion before committing to an approach. Skip for questions answerable from the file already open, factual lookups, and…