Skip to content

gillcash/necktie

v0.5.2MIT

An opinionated agent policy for decisions shaped by incentives, metrics, power, and hidden costs.

Necktie is an opinionated agent policy for decisions shaped by incentives, metrics, power, and extraction. It does not pretend every tradeoff is neutral.

Full is the useful default: it gives one Necktie judgment and completes or offers a concrete next action. Lite gives a focused Necktie judgment. Full can route accepted research-prompt work through a bounded prompt-reversal loop.

Choose the depth

ModeJudgment and actionBest fit
LiteStrongest accumulation and extraction case, followed by Necktie's rebuttalFocused decisions with a bounded judgment
FullLite plus an ambition pass and one context-specific useful actionDefault product, engineering, and strategy work

Necktie asks who benefits, who pays, who decides, who performs hidden labor, and who can leave. It distinguishes durable value creation from value capture, tests metrics for the behavior they reward, and looks for costs or risks that have been made invisible.

Its commitments are opinionated:

  • human agency over metric worship;
  • durable shared value over extraction;
  • consent, dignity, privacy, accessibility, security, and recourse;
  • truth over convenient narrative;
  • accountable power over opaque control.

Use Necktie

Necktie Core is active on every response through the host's native hook or instruction mechanism.

Invoke the explicit skill when you want a direct judgment:

/necktie We are considering ranking support agents by tickets closed per hour. Should we do it, and if so, how?

On skill-oriented hosts:

$necktie Audit this pricing plan. Who benefits, who pays, who controls the relationship, and who can leave?
$necktie Make the strongest case for controlling this market and identify the move with the highest expected leverage.

--mode lite|full is a one-shot skill override. It does not change session or configured defaults.

Full does the work already authorized. When a response would otherwise stop at an opinion, full offers one specific build or action. If deeper evidence is the next constraint, full will usually offer a portable research prompt. Invoke the prompt builder directly or accept the offer:

$necktie-research Reverse-engineer this discussion and the reference report into one copy-ready research prompt.

Necktie Research scans the user-authorized context, fingerprints any reference deliverable, critiques and reframes the inquiry, builds an exact prompt schema, reviews and revises the draft through a finite gate, and verifies that it works without hidden conversation state.

Necktie leads with a verdict or completed outcome, names the incentive or power imbalance that determined it, and recommends a concrete course.

Understand the plugin

The root plugin.json targets the Agent Plugins 1.0.0 specification. The portable surface contains the necktie judgment skill and the necktie-research prompt-building skill.

skills/necktie/references/policy.md is the canonical policy source. The build generates self-contained mode references plus one shared core/ preamble and mode deltas. Static rules inject Full. Dynamic hooks compose the selected mode.

necktie-mcp/ is an optional private stdio adapter. Its necktie prompt and read-only necktie_instructions tool accept a mode per request. MCP does not activate Necktie on every turn and exposes no arbitrary repository, file, execution, network, or mutation operation. The process is not a sandbox: it reads Necktie's bundled policy and optional local default configuration.

Install

Node.js must be available to hosts that run the lifecycle hook. Review third-party hooks before trusting them.

Claude Code

/plugin marketplace add gillcash/necktie
/plugin install necktie@necktie

Codex

codex plugin marketplace add gillcash/necktie
codex plugin add necktie@necktie

Open /hooks, review and trust the Necktie hooks, then start a new task. Restart the Codex desktop app after installation.

GitHub Copilot CLI

copilot plugin marketplace add gillcash/necktie
copilot plugin install necktie@necktie

Copilot namespaces the commands as /necktie:necktie and /necktie:necktie-mode.

Pi

pi install git:github.com/gillcash/necktie

OpenCode

Use the published package:

{ "plugin": ["@gillcash/necktie"] }

Or use a checkout:

{ "plugin": ["./.opencode/plugins/necktie.mjs"] }

Gemini CLI and Antigravity

gemini extensions install https://github.com/gillcash/necktie
agy plugin install https://github.com/gillcash/necktie

Hermes Agent

hermes plugins install gillcash/necktie --enable

Restart Hermes. It injects the selected policy before each model call and registers the necktie and necktie-mode commands. Use /necktie-mode to inspect or change the process-session mode.

Other supported hosts

HostInstallation or adapter
Devin CLIdevin plugins install gillcash/necktie
Grok Buildgrok plugin install gillcash/necktie --trust
Swivalswival skills add --global https://github.com/gillcash/necktie
OpenClawclawhub install necktie or copy .openclaw/skills/necktie/
Qoder.qoder/rules/necktie.md and optional hooks/qoder-hooks.json
Cursor.cursor/rules/necktie.mdc
Windsurf.windsurf/rules/necktie.md
Cline.clinerules/necktie.md
GitHub Copilot Chat.github/copilot-instructions.md
Kiro.kiro/steering/necktie.md
Aider, Zed, CodeWhale, Amp, JulesAGENTS.md
JunieAGENTS.md as the project guidelines file

See host support for adapter boundaries and installation checks.

Support and policies

Develop and validate

npm ci --prefix necktie-mcp
npm run build:adapters
npm test
python C:/Users/you/.codex/skills/.system/skill-creator/scripts/quick_validate.py skills/necktie
python C:/Users/you/.codex/skills/.system/skill-creator/scripts/quick_validate.py skills/necktie-research
python C:/Users/you/.codex/skills/.system/plugin-creator/scripts/validate_plugin.py .

skills/necktie/references/policy.md is the source for generated instruction artifacts and static adapters. Do not edit generated copies directly.

License

Necktie is available under the MIT License. Third-party attribution is recorded in NOTICE.