zero-hallucination-coder

Runs a disciplined Discuss -> Map -> Decompose -> Execute -> Verify loop that grounds code in verified structure — no invented APIs, no assumed imports, no placeholder code — with a lazy-senior-dev YAGNI ladder that deletes unnecessary code before it is written. Use when a coding task is high-stakes, complex, or spans existing code (auth, databases, migrations, multi-file features), or when the user explicitly asks to plan carefully before coding, avoid hallucinated code, or work rigorously. Not for trivial edits, typos, or throwaway one-off scripts — those do not need the full loop.

Quality

81%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

—

The risk profile of this skill

Quality

Content

73%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

A well-structured, highly actionable five-phase loop with excellent sequencing and validation checkpoints throughout. Its main weakness is length and some redundancy across the Anti-Patterns, Context Engineering Rules, and per-phase rules, which could be consolidated for better token efficiency.

Suggestions

Consolidate the Anti-Patterns (Rules 1-7) and Context Engineering Rules with the inline per-phase rules they duplicate, keeping one canonical statement of each rule to cut length without losing guidance.

Move the full template blocks (codebase map, story plan, verification report) and the Credits & Inspiration table into a reference file, leaving the phase sections as concise signposts.

Tighten the Ponytail ladder and right-sizing rule prose into compact checklists so the decision criteria scan at a glance.

Dimension	Reasoning	Score
Conciseness	Mostly efficient and free of basic-concept filler, but at ~270 lines it is ceremonially long — the Anti-Patterns list, Context Engineering Rules, and repeated inline phase rules overlap with material already embedded in each phase template, so several sections could be tightened.	3 / 5
Actionability	Provides concrete, copy-ready templates (codebase map, story plan, pre-check, self-review, handoff, verification report) with exact fields and decision rules (right-sizing thresholds, the six-rung Ponytail ladder), with only minor gaps where guidance is template scaffolding rather than executable steps.	4 / 5
Workflow Clarity	The five phases are clearly sequenced with explicit validation checkpoints at every boundary — 'Do not start planning until questions 1, 2, and 5 are answered,' 'stop if any UNKNOWN,' self-review verdicts, 'do not proceed until the user confirms,' and Phase 5 feedback loops that spawn micro-stories for gaps.	5 / 5
Progressive Disclosure	Single-file skill with well-organized sections and clearly signaled, one-level-deep cross-references to sibling skills (senior-architect, playwright-pro, self-improving-agent); some inline content (full template blocks, the credits table) could move to references, but structure and navigation are good.	4 / 5
	Total	16 / 20 Passed

Description

88%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

A strong, specific description that clearly conveys both the discipline loop and when to invoke it, with concrete triggers and anti-triggers. It could push trigger-term coverage to fully comprehensive with a few more synonyms, but otherwise lands near the top of the scale.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete actions — the five-phase loop (Discuss -> Map -> Decompose -> Execute -> Verify), 'no invented APIs, no assumed imports, no placeholder code,' and a YAGNI ladder that 'deletes unnecessary code before it is written' — giving comprehensive, specific coverage rather than vague language.	5 / 5
Completeness	Explicitly answers both what (the disciplined anti-hallucination loop plus YAGNI ladder) and when (high-stakes/complex/existing-code tasks, or explicit user requests), with concrete trigger phrases and an explicit anti-trigger ('Not for trivial edits, typos, or throwaway one-off scripts').	5 / 5
Trigger Term Quality	Includes natural user phrases ('plan carefully before coding,' 'avoid hallucinated code,' 'work rigorously') alongside domain triggers (auth, databases, migrations, multi-file features), though a few common synonyms are missing, keeping it just below the comprehensive anchor.	4 / 5
Distinctiveness Conflict Risk	Has a clear niche (zero-hallucination discipline loop) with explicit triggers and anti-triggers that reduce overlap, but it still borders on general careful-coding/planning skills, so minor conflict risk remains.	4 / 5
	Total	18 / 20 Passed

Validation

100%

Warnings & errors only

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: alirezarezvani/claude-skills
Path: engineering/zero-hallucination-coder/skills/zero-hallucination-coder/SKILL.md
Commit: 19392f7

Reviewed: ~3 hours ago

Table of Contents

Discovery Implementation Validation

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.