architect-spec

A procedure for writing a detailed implementation plan for a difficult compiler issue. It connects each reported pattern to the exact compiler code, explains the root cause, specifies the changes, and names tests to run.

In plain words
What is it for?
Use it to inspect an issue file, trace broken code-generation patterns, document file and function-level changes, record edge cases and risks, and select test262 tests for verification.
Why use it?
It turns a vague or complex bug report into an actionable plan before coding begins. This helps developers understand which functions emit the wrong WebAssembly instructions and how to verify the fix.

Skill for Claude CodeCodex

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/loopdive/js2/architect-spec
Any agent
npx skills add loopdive/js2 --skill architect-spec
Clone the repo
git clone --depth 1 https://github.com/loopdive/js2

Made for: Claude Code, Codex.

Per session 30 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 426 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 $0.00030 $0.00426
Opus 5 $0.00015 $0.00213
Sonnet 5 $0.00006 $0.00085
Haiku 4.5 $0.00003 $0.00043

Measured 2d ago against content hash 20b3191fe4b1, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

architect-spec 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 2d ago.

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.

.claude/skills/architect-spec/SKILL.md · 64 lines

What it actually says

Write Implementation Spec

Any agent can use this skill to produce an architect-quality implementation plan for an issue.

Steps

  1. Read the issue file at plan/issues/{N}.md

  2. For each sub-pattern in the issue, find the codegen function:

# Search for relevant function names
grep -n "function compile.*{keyword}" src/codegen/*.ts
# Or search for the Wasm instruction that's wrong
grep -n "{wasm_op}" src/codegen/*.ts
  1. Read the function and understand current behavior (what it emits and why it's wrong)

  2. Design the fix — what instructions should be emitted instead

  3. Write the spec in the issue file under ## Implementation Plan:

## Implementation Plan

### Root cause
[1-2 sentences]

### Work Item {letter}: {title} (~{N} CE)
**Patterns addressed**: {which sub-patterns from the breakdown}
**Risk**: {Low|Medium|High} — {why}
**Priority**: {1st|2nd|3rd}

#### Root cause
{specific function and line that emits wrong instructions}

#### Changes
**File: `src/codegen/{file}.ts`**
- Function `{name}` (line ~{N})
- {what to change and why}

#### Edge cases
- {case 1}
- {case 2}

#### Test
{specific test262 file to verify: before fix = CE, after fix = pass or runtime error}
  1. Recommend which work items to do first (highest impact, lowest risk)

Principles

  • Every fix should be ~10-20 lines, purely additive
  • No new repair passes — extend existing logic
  • No instruction stream restructuring
  • Each work item independently testable
  • Prefer compile-time fixes over post-hoc repairs
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. 2d ago First seen · 64 lines · 30 tokens per session scan A 20b3191fe4b1

Subscribe to this mod's changes

architect-spec is a skill published in the GitHub repository loopdive/js2 (59 stars, last pushed 2d ago), licensed Apache-2.0. It adds 30 tokens to every session and 426 once invoked, about $0.0002 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 2026-08-30.