New: 98,000+ skills indexed from GitHub repos with 100+ stars · every skill safety-checkedBrowse top skillsTop skills
Esc to closeBrowse all skills
Submit a skillBrowse skills
Submit a skillBrowse skills

Agent skill

Evidence First

by pmndrs in pmndrs/glyph

Shape human-facing engineering communication—including chat updates and final answers, reports, reviews, handoffs, debugging or benchmark summaries, PR and issue prose, READMEs, and technical…

MITAuto-check passedDevelopment

Install Evidence First

skills CLI
$ npx skills add pmndrs/glyph --skill evidence-first -a claude-code

Project install by default; add -g for ~/.claude/skills/.

GitHub CLI
$ gh skill install pmndrs/glyph evidence-first --agent claude-code

Project scope by default; add --scope user for a personal install. Needs GitHub CLI 2.90.0 or later (public preview).

Manual copy
$ git clone --depth 1 https://github.com/pmndrs/glyph.git skills-src && mkdir -p .claude/skills && cp -r skills-src/.agents/skills/evidence-first .claude/skills/evidence-first && rm -rf skills-src

Use ~/.claude/skills/ instead of .claude/skills for a personal install. The folder must contain SKILL.md.

Claude Code skills documentation · loads skills from .claude/skills/

Facts

Skill name
evidence-first
GitHub stars
395
Token cost
~1.2k tokens
SKILL.md length
579 words
Files
2
Skills in repo
8
Repo updated
First seen
Licence
MIT

At a glance

Shape human-facing engineering communication—including chat updates and final answers, reports, reviews, handoffs, debugging or benchmark summaries, PR and issue prose, READMEs, and technical…

  • Codex presents engineering work
  • Instructions only: no scripts, shell commands, URLs or credentials in SKILL.md
  • Findings to a human
  • Let document-specific frameworks retain their purpose and structure

What it does

Evidence First is an agent skill from pmndrs/glyph. Shape human-facing engineering communication—including chat updates and final answers, reports, reviews, handoffs, debugging or benchmark summaries, PR and issue prose, READMEs, and technical documentation—around clear claims, measured support, explicit gaps, and scoped decisions. Use whenever Codex presents engineering work, evidence, or findings to a human; let document-specific frameworks retain their purpose and structure.

Its SKILL.md is about 1.2k tokens, which your agent loads only when the skill is triggered. The skill folder holds 2 other files (for example `agents/openai.yaml`).

It sits in Development, covering Technical documentation. The repository describes itself as: ♠️ A typography engine for web graphics. The licence is MIT.

When your agent uses it

  • Codex presents engineering work
  • Findings to a human
  • Let document-specific frameworks retain their purpose and structure

Example prompts

  • “/evidence-first”

What it can do on your machine

Read from SKILL.md and the folder at commit b6ec800. It shows what the files ask for, not the result of running them.

  • Tool permissions

    Pre-approves nothing: there is no allowed-tools line, so your agent's usual permission prompts apply.

    From allowed-tools in the SKILL.md frontmatter.

  • Runs code

    No scripts in the folder and no shell commands in SKILL.md.

    From the folder's file list and the shell code blocks in SKILL.md.

  • Network

    No URLs in SKILL.md.

    From URLs in SKILL.md, links to its own repository left out.

  • Credentials

    Names no API keys, tokens, secrets or passwords.

    From names ending in _API_KEY, _TOKEN, _SECRET, _KEY or _PASSWORD in SKILL.md.

Context cost

Evidence First loads about 1.2k tokens when it runs. Until then it costs about 111 tokens; SKILL.md has 579 words of instructions outside code blocks.

Always · name and description, kept in context so the agent knows when to use it
~111
When it runs · the whole SKILL.md, loaded when a task matches
~1.2k

Estimates: characters ÷ 4, the usual rule of thumb; real counts depend on the model's tokenizer. Scripts and assets cost tokens only if the agent reads them.

Safety

Auto-check passed

The automated check found no risky patterns in SKILL.md.

Automated static check — not a guarantee. Review scripts before installing. It scans the text of SKILL.md for risky patterns (piping downloads into a shell, reading credential files, hidden Unicode, destructive commands); files beside SKILL.md are not scanned.

SKILL.md

The full file from pmndrs/glyph at commit b6ec800, republished under its MIT licence (© pmndrs). 579 words, ~1,153 tokens.

Download SKILL.mdSave it as .claude/skills/evidence-first/SKILL.md (or your agent's skills folder). This skill also uses 1 other file; get the full folder from GitHub.
name
evidence-first
description
Shape human-facing engineering communication—including chat updates and final answers, reports, reviews, handoffs, debugging or benchmark summaries, PR and issue prose, READMEs, and technical documentation—around clear claims, measured support, explicit gaps, and scoped decisions. Use whenever Codex presents engineering work, evidence, or findings to a human; let document-specific frameworks retain their purpose and structure.

Working orientation

Reason freely internally. At the human boundary, translate that exploration into the clearest useful account of the work: usually a claim, the support that matters, and any uncertainty that could change a decision. This is an orientation for judgment, not a required response shape.

Let the situation suggest the shape

Treat these as useful signals rather than rules:

  • A short chat answer often wants direct prose: the outcome first, followed by the evidence or gap that changes the reader's understanding. Headings may add more ceremony than clarity.
  • Ongoing work is often easiest to follow when the update reveals the parts that materially changed: perhaps what is known now, what remains uncertain, or where the work is heading.
  • A longer report, handoff, debugging summary, benchmark report, or PR description may benefit from answering selected reader questions below.
  • A code or design review is usually more useful when actionable findings appear first and are ordered by impact; a shipped-work narrative may be irrelevant.
  • A document-specific framework should normally choose the document's purpose and structure, with evidence-first habits operating inside that shape.

Depart from these defaults whenever audience, medium, or task calls for a clearer form.

Prefer the smallest useful artifact

Simple claims often need only prose. Code, a table, a measurement block, or a diagram earns its place when it materially clarifies behavior, comparison, structure, or evidence—not merely as an alternative to writing a paragraph.

Reader questions for report-like outputs

Longer engineering artifacts often become clearer when they answer the relevant questions below. They are prompts for judgment, not headings to reproduce or a preferred ordering.

  • What changed or was learned?
  • What observation, command result, measurement, or artifact supports the important claim?
  • Did the work reveal a material finding outside the requested scope?
  • What meaningful gap was not verified?
  • Does a decision require the reader's authority, and what trade-off makes it theirs?
  • If work remains, what next action has the highest value?
Show full SKILL.md (257 more words)Show less

Evidence habits

Use these habits to improve trust without turning every response into an audit:

  • Make observed, inferred, and not verified distinguishable when the difference matters.
  • Strong mechanism claims deserve supporting evidence. When the cause remains uncertain, a hypothesis and a useful disconfirming check are often more honest and actionable.
  • Before attributing a failure, consider whether both the probe and the product have been tested at the lowest honest layer.
  • Corrections are clearest when stated plainly and carried forward without ceremony.
  • Reversible, in-scope decisions usually benefit from autonomous progress. Destructive, external, costly, or materially scope-expanding decisions usually benefit from being surfaced.
  • A question earns the interruption when its answer can materially change the result or scope.
  • When shortening an output, preserve decisions, caveats, material evidence, and required facts before background or repetition.

Layer documentation frameworks

Other skills retain authority over their subject matter. A technical or review skill determines what work to perform and what counts as valid evidence; Evidence First helps translate the result for a human.

Open Knowledge Format can own bundle structure, provenance, lifecycle, and navigation. Evidence First can shape the claims and uncertainty inside each concept without adding fields or changing conformance rules.

When a documentation framework applies, it can own document purpose and top-level structure while evidence-first habits shape claim quality, evidence, uncertainty, and decision boundaries inside it.

For Diátaxis specifically:

  • keep tutorials learning-oriented;
  • keep how-to guides goal-oriented;
  • keep reference precise and scannable;
  • keep explanations understanding-oriented.

These report-oriented questions are therefore usually unnecessary for those document types and for short conversational answers.

© pmndrs, MIT. Rendered from Markdown: HTML in the file is shown as text, images as links, and headings moved down two levels. Raw file

Files

SKILL.md and 1 other file in .agents/skills/evidence-first of pmndrs/glyph.

  • SKILL.md
  • agents/openai.yaml

Open the folder on GitHubat commit b6ec800

Compare with similar skills

Evidence First next to the 5 skills that share the most tags, products or categories with it. Stars are the repository's; “used in” counts other GitHub owners with a copy.

Evidence First compared with similar skills
SkillStarsUsed inTokensAuto-checkLicenceRepo updated
Evidence First this skillpmndrs/glyph395—~1.2kAutomated safety check: PassMIT
Diagram Designcathrynlavery/diagram-design47k1 repos~7.5kAutomated safety check: PassMIT
Simple Englishmoeru-ai/airi50k2 repos~4.6kAutomated safety check: PassMIT
Doc SyncJetBrains/ideavim10k2 repos~2.6kAutomated safety check: PassMIT
Mailspring App ScreenshotsFoundry376/Mailspring18k—~1.5kAutomated safety check: PassGPL-3.0
Draw.io Diagram StudioAgents365-ai/drawio-skill10k—~2.4kAutomated safety check: NotesMIT

Similar skills

  • Diagram Design

    cathrynlavery/diagram-design

    Creates branded diagrams, from architecture, flowchart and sequence to charts and maps, as self-contained HTML with inline SVG, with import from draw.io, Mermaid and Excalidraw.

    47k GitHub starsUsed in 1 repo~7.5k tokens
    DevelopmentAuto-check passed

  • Simple English

    moeru-ai/airi

    Write or rewrite technical text with the rules of ASD-STE100 Simplified Technical English so it is clear, unambiguous, and free of AI slop.

    50k GitHub starsUsed in 2 repos~4.6k tokens
    DevelopmentAuto-check passed

  • Doc Sync

    JetBrains/ideavim

    Official

    Keeps IdeaVim documentation in sync with code changes. An agent skill from JetBrains/ideavim.

    10k GitHub starsUsed in 2 repos~2.6k tokens
    DevelopmentAuto-check passed

  • Mailspring App Screenshots

    Foundry376/Mailspring

    Captures screenshots of the running Mailspring dev app for docs, PRs or visual checks by launching it with a debugging port, driving the UI and clipping to an element.

    18k GitHub stars~1.5k tokensUpdated today
    DevelopmentAuto-check passed

  • Draw.io Diagram Studio

    Agents365-ai/drawio-skill

    Creates and edits editable draw.io diagrams from descriptions, code, infrastructure files, SQL and API schemas, with sync, review, test and export tools.

    10k GitHub stars~2.4k tokensUpdated 7 days ago
    DevelopmentAuto-check: notes

  • Dark Architecture Diagram Builder

    Cocoon-AI/architecture-diagram-generator

    Creates dark-themed system, cloud, security and network architecture diagrams as self-contained HTML files with inline SVG and CSS.

    7.4k GitHub starsUsed in 1 repo~2.1k tokens
    DevelopmentAuto-check passed

More from pmndrs/glyph

All 8 skills in this repo

  • Codemod

    pmndrs/glyph

    Author, archive, apply, and verify TypeScript codemods with ts-morph for APIs that have reached the remote default branch or external users.

    395 GitHub stars~1.2k tokensUpdated today
    Auto-check passed

  • Open Knowledge Format

    pmndrs/glyph

    Create, migrate, inspect, query, validate, or maintain Open Knowledge Format v0.2 bundles made from linked Markdown concepts with YAML provenance.

    395 GitHub stars~2.5k tokensUpdated today
    Auto-check passed

  • Tsl

    pmndrs/glyph

    Implement, migrate, review, debug, or verify Three.js Shading Language (TSL) materials, node graphs, WebGPU compute work, and post-processing.

    395 GitHub stars~1.5k tokensUpdated today
    Auto-check passed

  • Diataxis Docs

    pmndrs/glyph

    Design, classify, write, audit, or restructure technical documentation with the Diátaxis framework.

    395 GitHub stars~1.4k tokensUpdated today
    Auto-check passed

  • Maintainability Review

    pmndrs/glyph

    Audit and improve repository code milestone by milestone for correctness, clarity, local reasoning, DRY design, explicit state modeling, panic resistance, and trustworthy TypeScript boundaries.

    395 GitHub stars~1.7k tokensUpdated today
    Auto-check passed

  • Gh Stack

    pmndrs/glyph

    Manage dependent branches and pull requests with the gh-stack GitHub CLI extension.

    395 GitHub stars~1.4k tokensUpdated today
    Auto-check passed

Categories

Questions about Evidence First

What does Evidence First do?

Shape human-facing engineering communication—including chat updates and final answers, reports, reviews, handoffs, debugging or benchmark summaries, PR and issue prose, READMEs, and technical…. Evidence First is an agent skill from pmndrs/glyph. Shape human-facing engineering communication—including chat updates and final answers, reports, reviews, handoffs, debugging or benchmark summaries, PR and issue prose, READMEs, and technical documentation—around clear claims, measured support, explicit gaps, and scoped decisions.

When should I use Evidence First?

Evidence First fits situations like: Codex presents engineering work; findings to a human; let document-specific frameworks retain their purpose and structure.

How do I install Evidence First in Claude Code?

Run `npx skills add pmndrs/glyph --skill evidence-first -a claude-code`. Or copy the skill folder (.agents/skills/evidence-first in pmndrs/glyph) into .claude/skills/evidence-first in your project. Claude Code loads it when a task matches its description.

How do I install Evidence First in Codex?

Run `npx skills add pmndrs/glyph --skill evidence-first -a codex`. Or copy the skill folder (.agents/skills/evidence-first in pmndrs/glyph) into .agents/skills/evidence-first in your project. Codex loads it when a task matches its description.

Can I use Evidence First in Cursor, Gemini CLI or GitHub Copilot?

Cursor, Gemini CLI, GitHub Copilot and OpenCode also load SKILL.md folders. With the skills CLI, run `npx skills add pmndrs/glyph --skill evidence-first -a cursor` (or -a gemini-cli, github-copilot or opencode for the others). To copy it by hand, put the folder in .cursor/skills/evidence-first, .gemini/skills/evidence-first, .github/skills/evidence-first and .opencode/skills/evidence-first in your project.

What does Evidence First need to run?

SKILL.md names no scripts, command-line tools or credentials: Evidence First is instructions for the agent only.

Does Evidence First access the network?

SKILL.md contains no URLs. Any network use would come from the scripts or tools the agent runs. This is read from the text; nothing was executed.

Is Evidence First safe to install?

Our automated static check of SKILL.md found no risky patterns, such as piping downloads into a shell, reading credential files or hidden Unicode. It is not a guarantee. Review the folder before installing.

What licence does Evidence First use?

Evidence First is published under the MIT licence (the repository's licence). It allows redistribution, so the full SKILL.md is shown on this page.

How many tokens does Evidence First use?

About 1.2k tokens (SKILL.md is roughly 4.6k characters). Agents keep only the skill's name and description in context until a task matches; then they load SKILL.md in full.

What are the alternatives to Evidence First?

Skills that share tags, products or a category with Evidence First: Diagram Design (cathrynlavery/diagram-design, 47k stars), Simple English (moeru-ai/airi, 50k stars), Doc Sync (JetBrains/ideavim, 10k stars) and Mailspring App Screenshots (Foundry376/Mailspring, 18k stars). The comparison table on this page puts their stars, adoption, token cost, safety result and licence side by side.

Who maintains Evidence First?

pmndrs (a GitHub organization) maintains it in pmndrs/glyph, which has 395 GitHub stars. The repository holds 8 skills in this directory. The repository was last updated on October 9, 2026.

Source: pmndrs/glyph on GitHub. Facts on this page come from the repository at the commit we read; the author's words are quoted as theirs.