tp-design-complete

tp-design-complete is a skill for Claude Code, Codex from CurtisThe/three-pillars-plugin. It costs 52 tokens per session (7,792 once invoked), scanned A, original, Apache-2.0.

A review tool that checks a detailed software design against the existing codebase before implementation.

In plain words
What is it for?
Use it to inspect a design's assumptions against the repository and review the design from multiple technical angles.
Why use it?
It can reveal mismatched interfaces, conflicting data structures, resource limits, and other feasibility problems before code is written.

Skill for Claude CodeCodex

Part of the three-pillars plugin — 37 skills, 20 agents shipped together

Install

Getting it into your agent

One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.

agentmods
npx agentmods add skills/curtisthe/three-pillars-plugin/tp-design-complete
Any agent
npx skills add CurtisThe/three-pillars-plugin --skill tp-design-complete
Clone the repo
git clone --depth 1 https://github.com/CurtisThe/three-pillars-plugin

Made for: Claude Code, Codex.

Or install three-pillars, the plugin that ships this one along with the rest of its 37 skills, 20 agents.

Wrote this? Show the measurements

A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.

agentmods badge for tp-design-complete

README.md
[![agentmods](https://agentmods.dev/badge/skills/curtisthe/three-pillars-plugin/tp-design-complete.svg)](https://agentmods.dev/skills/curtisthe/three-pillars-plugin/tp-design-complete)
Your own site
<a href="https://agentmods.dev/skills/curtisthe/three-pillars-plugin/tp-design-complete"><img src="https://agentmods.dev/badge/skills/curtisthe/three-pillars-plugin/tp-design-complete.svg" alt="Measured on agentmods" height="20"></a>
Per session 52 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 7,792 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
Origin original No closer match found in the catalogue.
Token cost

What it costs to keep this loaded

Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.

ModelPer sessionOnce invoked
Fable 5 $0.00052 $0.07792
Opus 5 $0.00026 $0.03896
Sonnet 5 $0.00010 $0.01558
Haiku 4.5 $0.00005 $0.00779

Measured 5d ago against content hash c5812f954737, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

tp-design-complete scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 5d ago.

The scan reads SKILL.md. This mod also ships 7 executable files (scripts/detect_parent.py, scripts/test_design_complete_archive_guard.py, scripts/test_design_complete_skill_md.py, …), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.

Nothing flagged

None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.

skills/tp-design-complete/SKILL.md · 215 lines

How it starts

The opening of the file, as written. The whole thing — 215 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Design Complete

Mark a finished design as complete, archive it out of the active designs directory, and clean up session state.

Argument: {design-name} (required) — must match an existing directory under three-pillars-docs/tp-designs/.

Steps

  1. Run first-run preflight per skills/_shared/first-run.md.

0a. Run cwd preflight per skills/_shared/cwd-preflight.md: python3 "$TP_ROOT"/skills/_shared/cwd_preflight.py {design-name}. Exit 3 → stop and show the cd fix. Exit 0 → continue.

  1. Validate the design name per skills/_shared/validate-name.md.
  2. Resolve the design directory: three-pillars-docs/tp-designs/{design-name}/. If it doesn't exist, tell the user and stop.
  3. Check if -learn has been run (hard-blocks unless --skip-learn was passed): Determine whether this is a spike (has spike-results.md) or a full design (has implementation-audit.md).
    • For spikes: Check if the product_roadmap.md Design Inventory table has been updated with the spike's verdict (GO/PARTIAL/NO-GO). If not, refuse to proceed:

      Refusing to complete: /tp-spike-learn hasn't been run yet — the Design Inventory doesn't carry a GO/PARTIAL/NO-GO verdict. That step propagates findings into project docs and scans for affected downstream designs; skipping it leaves the roadmap silently stale. Run /tp-spike-learn {design-name} first, then re-invoke this skill. If you have a specific reason to skip (e.g., the spike was a one-off scratchpad with no roadmap presence), re-invoke with --skip-learn.

    • For full designs: Check if the product_roadmap.md Design Inventory status reflects completion. If it still says "Designed" or "Pending", refuse to proceed:

      Refusing to complete: /tp-design-learn hasn't been run yet — the Design Inventory status is still "Designed" / "Pending" / "Implementing". That step propagates implementation results into product_roadmap.md, architecture.md, and known_issues.md, updates the Design Inventory status, and scans for affected sibling designs. Skipping it causes the roadmap to drift silently across the completion boundary. Run /tp-design-learn {design-name} first, then re-invoke this skill. If you have a specific reason to skip — legacy designs that pre-date this rule, or designs where learn was run out-of-band and the roadmap status is intentionally stale — re-invoke with --skip-learn.

    • --skip-learn escape hatch: If the user passed --skip-learn, log a one-line notice (Proceeding without learn per --skip-learn flag) and continue. The flag exists for legacy designs and well-understood out-of-band cases; it must not become the default workflow. Do not silently treat absence-of-flag as override.
  4. Show a summary of what will happen:
    • The completion date that will be stamped (today's date, YYYY-MM-DD)
    • Whether handoff.md exists and will be archived (banner added)
    • The destination: three-pillars-docs/completed-tp-designs/{design-name}/
    • The current branch and the base branch — after archival, a commit will be made on the current branch and (if the current branch is not the base) a PR back to base will be offered
  5. Ask for confirmation before proceeding. A simple "Complete this design?" is enough.
  6. On confirmation, execute these steps in order:

Read the full file on GitHub · 215 lines

Changes

What this file has done since we first saw it

Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.

  1. 5d ago First seen · 215 lines · 52 tokens per session scan A c5812f954737

Subscribe to this mod's changes

tp-design-complete is a skill published in the GitHub repository CurtisThe/three-pillars-plugin (4 stars, last pushed 1mo ago), licensed Apache-2.0. It adds 52 tokens to every session and 7,792 once invoked, about $0.0003 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.