Collected from public GitHub. Every version is saved exactly as it was and never changes. Install it at an exact version with the mdr tool, and every file is checked before it reaches your agent. The main file is checked against the SHA-256 fingerprint recorded here, and the others against the fingerprints GitHub keeps for its source commit. The automatic checks below grade the latest version, and the same checks always give the same file the same grade.
What the file says
# Why
> Ported from the `why` skill in [cursor/plugins pstack](https://github.com/cursor/plugins/tree/main/pstack/skills) — Cursor, a separate AI-coding-editor product, not to be confused with a text/DB
> cursor — for use with this harness's `Agent` tool. Agent-spawning mechanics use this repo's actual subagent types and model/effort params in place of the original's source-specific model names and
> `readonly` flag; MCP discovery is adapted to how this session surfaces connected servers.
>
> **Git-history steps prefer a code-graph MCP first, when one is connected.** If your environment has one (e.g. `tokensave`), prefer its graph tools over raw shell git commands for codebase research
> wherever an equivalent exists — a per-symbol blame/log tool gives structural history tracked across renames and is the first move for any target that's a function, method, class, or type. Raw
> `git blame`/`git log --follow -p` are the fallback for what a code-graph tool doesn't cover (or for when none is connected): line-precise blame that doesn't align to a symbol boundary, full raw
…
The line at the top of this page pins the label git:20260824.34ad58b, which is what most people want. This one pins the sha256 fingerprint of the contents instead, so it survives a label being moved. Either way the exact version is written to mdr.lock, and mdr install downloads and checks those contents again, so it installs them exactly or stops with an error.
B 16 of 17 checks passed. Fixed automatic checks with no AI involved, so the same checks give the same answer every run.
fail: Frontmatter declares a description
pass: Frontmatter block present
pass: Frontmatter declares a name
pass: Size between 200 bytes and 200 KB (28802 bytes)
pass: No zero-width or bidi control characters
pass: No instruction hidden inside an HTML comment
pass: No link to an exfiltration or paste host
pass: No credential-shaped string
pass: No instruction to send local credentials anywhere
pass: No text hidden with inline styles
pass: No prompt-injection phrasing
pass: No curl or wget piped into a shell
pass: No recursive delete of root, home or parent
pass: No instruction to read or print local credentials
pass: No base64 blob over 200 characters
pass: No link to a raw IP address
pass: No script tag
Starting a new one? Use a template for SKILL.md on mdpantry, which passes every one of these checks and says what to change in it.
Source
GitHub
pantheon-org/tekhne · 10 stars · license MIT · pushed 2026-10-21 · branch main
API
GET https://markdownregistry.com/api/v1/artifacts/art_a7gb55l3zlp6yvob
GET https://markdownregistry.com/api/v1/resolve?ref=pantheon-org/tekhne/why
GET https://markdownregistry.com/api/v1/blob/996b3e7243fa49bb49f25c839d6a34943a745d7b976899bbc21f05e2bfb6d000
Your AI assistant does the legwork. You hear about the deals worth your word. Hand yours the standing instructions at modelranch.com and it joins the network that reads files like this one.
pantheon-org/tekhne · skills/agentic-harness/cache-audit/SKILL.md · Audit your Claude Code setup against prompt caching best practices. Checks ordering, tool stability, dynamic content…
pantheon-org/tekhne · skills/agentic-harness/create-context/SKILL.md · Create the baseline session context from the in-folder of raw project documents, classifying each file by priority…
pantheon-org/tekhne · skills/agentic-harness/guided-interview/SKILL.md · Conduct a structured, one-question-at-a-time interview to explore a specific topic or decision — each question offers…
pantheon-org/tekhne · skills/agentic-harness/load-context/SKILL.md · Resume session from CONTEXT-llm.md. Use when resuming work, loading saved context, continuing a previous session…
pantheon-org/tekhne · skills/agentic-harness/opencode-toolkit/build-plugins/SKILL.md · Create OpenCode plugins using the opencode-ai/plugin SDK. Use when user wants to build a plugin, extend OpenCode…
pantheon-org/tekhne · skills/agentic-harness/opencode-toolkit/build-tool/SKILL.md · Build custom tools and plugins using the OpenCode SDK packages. Use when creating custom tools, implementing tool…
pantheon-org/tekhne · skills/agentic-harness/opencode-toolkit/configure/SKILL.md · Configure OpenCode via opencode.json and AGENTS.md with deterministic provider setup, model selection, permission…
cursor/plugins · harness/pstack/claude-code/skills/why/SKILL.md · Use for 'why does X work this way', 'why we picked Y', design rationale, regressions, postmortems, or data-backed…
cursor/plugins · pstack/skills/why/SKILL.md · Use for 'why does X work this way', 'why we picked Y', design rationale, regressions, postmortems, or data-backed…
michael-denyer/pstack-claude · plugins/pstack/skills/why/SKILL.md · Use for 'why does X work this way', 'why we picked Y', design rationale, regressions, postmortems, or data-backed…
backnotprop/pstack · skills/why/SKILL.md · Use for 'why does X work this way', 'why we picked Y', design rationale, regressions, postmortems, or data-backed…
bastani-inc/atomic · packages/workflows/skills/why/SKILL.md · Use for 'why does X work this way', 'why we picked Y', design rationale, regressions, postmortems, or data-backed…
millw14/merrymen · plugins/merrymen/skills/why/SKILL.md · Explain why the user's Merrymen trading agent has or hasn't traded, from its recorded state, blockers and recent…
j-star-films-studios/vibecode-protocol-suite · assets/.agent/skills/code-intelligence/why/SKILL.md · Use for 'why does X work this way', 'why we picked Y', design rationale, regressions, postmortems, or data-backed…
creedants/pstack-t3 · t3/overrides/why/SKILL.md · Use for 'why does X work this way', 'why we picked Y', design rationale, regressions, postmortems, or data-backed…
