Skip to content
Agentmart
Submit a tool
This entry duplicates niko-preflight, which is the one we list in search.

niko-preflight

Niko Memory Bank System - Preflight Phase - Pre-Build Plan Validation

Skill by Texarkanine

IndexedAGPL-3.0
Write a review

Chalk Labs

Need help deploying niko-preflight?

Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.

Book a call

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/texarkanine-niko-preflight

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

Niko Memory Bank System - Preflight Phase - Pre-Build Plan Validation

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions: not checked yet.

What's inside

Skill name
niko-preflight
Description
Niko Memory Bank System - Preflight Phase - Pre-Build Plan Validation
License
AGPL-3.0
Instructions
105 lines in SKILL.md · view on GitHub

Files (1)

  • SKILL.md

SKILL.md (first lines)

# Preflight Phase - Pre-Build Plan Validation

This command validates the implementation plan against codebase reality before any code is written. It catches design oversights, convention conflicts, TDD violations, and integration issues that would otherwise surface during or after the build.

## Step 1: Load Memory Bank Files

Read:
- `memory-bank/active/tasks.md`
- `memory-bank/active/projectbrief.md`
- `memory-bank/systemPatterns.md`
- `memory-bank/techContext.md`
- `memory-bank/active/creative/**/*.md` (if any exist)

## Step 2: Preflight Workflow

1. **Verify Prerequisites**
   - Check `memory-bank/active/tasks.md` for planning completion
   - For Level 3-4: Verify creative phase documents exist (if creative phases were flagged)
   - Read implementation plan and design decisions

2. **TDD Plan Encoding** *(blocking)*
   - The test-first process lives in `.cursor/rules/shared/always-tdd.mdc`
   - This check governs units that change executable behavior. A unit delivering user-facing prose or policy (docs content, PR/issue templates, CONTRIBUTING, instructional comments, rule/skill wording, etc.) owes no tests for those artifacts; omitting tests for those artifacts passes this check
   - For each implementable unit of executable work (function, slice, milestone — whatever granularity the plan uses), confirm the ordered substeps place test-writing before production code, explicitly enough that a reasonable implementer cannot follow the plan by coding first
   - When a numbered step is a scheduled change-detector (a test that can only go red when someone deliberately edits the artifact it asserts on — heading, phrase, link, or checklist assertions on a document), delete that step. Keep the other steps. Record the finding and continue.
   - When a unit already has both test steps and production steps and they are in the wrong order, put the test steps first. Same steps. Record the finding and continue.
   - Do not invent tests. Do not emit always-tdd stages.
   - FAIL when the numbered steps for an executable unit have no test steps (implementation-only under a "we follow TDD" disclaimer, or TDD only in the preamble). This still applies after a change-detector strike.
   - On FAIL: cite the executable units lacking test steps. Write `FAIL (blocking)`.

3. **Convention Compliance**
   - Verify the plan's proposed file locations, naming conventions, and patterns align with established codebase conventions documented in `memory-bank/systemPatterns.md`
   - Cross-reference proposed module structure against existing project organization
   - Flag any deviation from established patterns with specific recommendations

4. **Dependency Impact**
   - Trace the plan's touchpoints through the dependency graph
   - Identify modules, consumers, or tests that will be affected but aren't accounted for in the plan
   - Verify that all downstream impacts are documented and addressed

5. **Conflict Detection**
   - Search for existing implementations, utilities, or patterns that overlap with or contradict the plan's approach
   - Identify duplication-in-waiting - cases where the plan proposes building something the codebase already provides
   - Flag any proposed changes that would break public contracts or published interfaces — internal restructuring that preserves the public API surface is not a conflict

6. **Completeness Precheck**
   - Verify the plan addresses all stated requirements with concrete implementation steps mapped to each one - not aspirationally, but with specific files, functions, and approaches identified
   - Flag any requirements that are acknowledged but lack a clear implementation path
   - Verify test coverage is planned for all new executable behavior — not for prose or policy artifacts; the TDD Plan Encoding check governs that boundary

7. **Radical Innovation** *(advisory - not blocking)*
    - What's the single smartest and most radically innovative and accretive and useful and compelling change you could make to the plan at this point?
    - Describe the change concretely - not as a vague suggestion, but as a specific structural sketch the operator can evaluate against the cost of redesign.
    - Record that idea as an advisory finding. Do not make the change to the plan, even if the idea fits the brief.

8. **Judge, Do Not Fix**
   - Surface and judge. Never modify the plan under review, except the TDD step swap and change-detector strike above.
   - Allowed writes only: `memory-bank/active/.preflight-status`, the `**Phase:**` field in `activeContext.md` (under **End of Verification**), `progress.md`, and those two in-phase plan edits on `tasks.md`.
   - Do not rewrite Implementation Plan units, behavior lists, or other scheduled work except that swap and that strike.
   - Record every issue as a finding. FAIL when the plan must change before build (`FAIL (fixable)` or `FAIL (blocking)`); PASS only when the plan is acceptable as-is (advisories allowed).

9. **Write Status**
   - Overwrite `memory-bank/active/.preflight-status`. First line is exactly one allowed value from `.cursor/rules/shared/niko/memory-bank/active/preflight-status.mdc`. After a blank line, write this run's findings.

## Step 3: Log Progress

> 🚨 **Printing this notice is NOT the end of this phase.** After printing, continue immediately to the next step - do not stop.

Update `memory-bank/active/progress.md` to record that Preflight completed and what the first line of `.preflight-status` was.

Print the appropriate block:

### PASS

~~~markdown
# Preflight Result

✅ PASS

## Findings

1. **Findings** - bulleted list of each finding with severity
2. **Advisory items** (if any) - concrete recommendations the operator can evaluate

~~~

### FAIL

~~~markdown
# Preflight Result

❌ FAIL

## Findings

1. **Findings** - bulleted list of each finding with severity
2. **Advisory items** (if any) - concrete recommendations the operator can evaluate

~~~

## Step 4: End of Verification

Update `memory-bank/active/activeContext.md` so `**Phase:**` records Preflight complete with the first line of `.preflight-status` (e.g. `**Phase:** PREFLIGHT - COMPLETE (PASS)`). Do not load a level workflow or begin another phase. Stop.

From Texarkanine/open-plugins-surface-test, AGPL-3.0.

Signals

Stars
0
Forks
0
Open issues
0
Last commit (this folder)
Sep 2, 2026 (2 months ago)

From GitHub (fetched Oct 24, 2026).

Details

License
AGPL-3.0
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

Same skill elsewhere

Copies of this SKILL.md published by others; search shows only one of them.

More from Texarkanine

Done for you

Rather not set it up yourself?

Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.

See how we build AI agents →Read the case studies

Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.

launches shipped
50+
launches shipped
media placements
500+
media placements
founders on every account
2
founders on every account

Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block

Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.

Get niko-preflight running in your business

Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.

Goes to the Chalk Labs founders, not the newsletter. Privacy

Need help deploying niko-preflight?

CallGet help
Cite this page

niko-preflight skill. Agent Mart. https://agentmart.dev/skills/texarkanine-niko-preflight (last updated 2026-10-24)