evidence-based-coding

Ground code, configuration, and completion claims in evidence for the current candidate and relevant environment.

IndexedMIT

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/leo110047-evidence-based-coding

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.

About

Ground code, configuration, and completion claims in evidence for the current candidate and relevant environment. Use when inspecting, changing, or reporting repository behavior.

Use cases: agent-building

Reviews

Write a review

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

Permissions

What this skill can do and what it asks for, from its SKILL.md. Checked Oct 25, 2026.

Tools: No tools found
Secrets: None asked for
Install scripts: None found

Network hosts it talks to 1

codeload.github.com · found in install command

Files it touches

~/.claude/skills/evidence-based-coding

What's inside

Skill name: evidence-based-coding
Description: Ground code, configuration, and completion claims in evidence for the current candidate and relevant environment. Use when inspecting, changing, or reporting repository behavior.
License: MIT
Instructions: 26 lines in SKILL.md · view on GitHub

Agent Skills spec issues

allowed-tools: must be a space-separated string

Files (5)

SKILL.md
completion-verification.md · reference/
goal-verification.md · reference/
hallucination-patterns.md · reference/
verification-workflows.md · reference/

SKILL.md (first lines)

# Evidence-Based Coding

Check the authoritative code, configuration, output, or logs before making a
claim. Match the evidence to the scope and boundary being claimed.

- Read relevant context behind search results and agent reports.
- Static inspection proves structure; mocks and type checks do not establish
  live provider, authorization, process, platform, or deployment behavior.
- Evidence may be reused when it still covers the same candidate, inputs,
  relevant environment, and requirement. A new conversational turn alone does
  not invalidate it. Recheck state that may have changed.
- Run checks required by the repository and the changed risk. Repeat or broaden
  them only after relevant changes, failures, or unresolved concerns.
- Never weaken, skip, or delete an assertion, test, type, or lint rule to pass.
  Correct a wrong gate together with its stated policy and regression coverage.
- Report unavailable verification, partial results, and assumptions explicitly.
  Do not turn a plausible result or another agent's summary into a verified fact.

Load only the reference needed:

- `reference/completion-verification.md`: evidence reuse and completion.
- `reference/verification-workflows.md`: choosing proof for a claim.
- `reference/hallucination-patterns.md`: misleading evidence patterns.
- `reference/goal-verification.md`: tracing a requested outcome to behavior.

From leo110047/goldband, MIT.

Signals

Stars: 0
Forks: 0
Open issues: 5
Last commit (this folder): Oct 14, 2026 (15 days ago)

From GitHub (fetched Oct 24, 2026).

Details

License: MIT
Repository: github.com/leo110047/goldband
First seen: Oct 24, 2026
Last checked: Oct 24, 2026
Source: GitHub github-graphql

More from leo110047

Cite this page

evidence-based-coding skill. Agent Mart. https://agentmart.dev/skills/leo110047-evidence-based-coding (last updated 2026-10-24)