Pp GitHub Contents

<!-- GENERATED FILE — DO NOT EDIT.
     This file is a verbatim mirror of library/developer-tools/github-contents/SKILL.md,
     regenerated post-merge by tools/generate-skills/. Hand-edits here are
     silently overwritten on the next regen. Edit the library/ source instead.
     See the repository agent guide, section "Generated artifacts: registry.json, cli-skills/". -->

GitHub Contents — Printing Press CLI

Prerequisites: Install the CLI

This skill drives the github-contents-pp-cli binary. You must verify the CLI is installed before invoking any command from this skill. If it is missing, install it first:

1. Install via the Printing Press installer. It defaults binaries to $HOME/.local/bin on macOS/Linux and %LOCALAPPDATA%\Programs\PrintingPress\bin on Windows:

   npx -y @mvanhorn/printing-press-library install github-contents --cli-only

2. Verify: github-contents-pp-cli --version
3. Ensure the reported install directory is on $PATH for the agent/runtime that will invoke this skill.

If the npx install fails (no Node, offline, etc.), fall back to a direct Go install (requires Go 1.26.6 or newer). This installs into $GOPATH/bin (default $HOME/go/bin), so add that directory to $PATH instead:

go install github.com/mvanhorn/printing-press-library/library/developer-tools/github-contents/cmd/github-contents-pp-cli@latest

If --version reports "command not found" after install, the runtime cannot see the binary directory on $PATH. Do not proceed with skill commands until verification succeeds.

Addresses owner/repo/path#ref like degit, but streams big files past the 1 MB contents-API limit via the raw CDN, preserves folder structure, and gives every command --json for agents. The plan/verify/sync-dir loop — preview cost before downloading, prove local files match remote by git blob SHA, then fetch only the diff — exists in no other GitHub download tool.

When to Use This CLI

Reach for this CLI when a task needs files out of a GitHub repo without a full clone: downloading a subdirectory (docs, datasets, templates, book collections) with structure preserved, previewing what a download costs, verifying a local copy still matches upstream, or keeping a mirror current with minimal bandwidth. It works unauthenticated on public repos and uses at most a couple of API requests per operation.

Anti-triggers

Do not use this CLI for:

• Cloning a repo with git history — use git clone or gh repo clone
• Working with issues, pull requests, or actions — use the gh CLI
• Writing, committing, or pushing files to a repo — this CLI is read-only
• Downloading Git LFS binaries — v1 detects and warns but does not resolve LFS pointers
• Searching code across GitHub — use gh search or the code-search API

Unique Capabilities

These capabilities aren't available in any other tool for this API.

Plan-verify-sync loop

• plan — Preview exactly what a fetch would download — file list, sizes, total bytes, API-request cost vs your remaining quota, and LFS-pointer warnings — before spending any bandwidth.

  Run this before fetch to know a download is 122 files / 1.9 GB and costs only 2 API requests, instead of finding out mid-download.

  github-contents-pp-cli plan mjwoon/AI-readings/books --agent

• verify — Check whether a previously downloaded directory still matches the remote at a ref — match/changed/missing/extra per file — without re-downloading anything.

  Proves 'this workspace matches ref X' cheaply — the post-download assertion CI and agents otherwise can't make without re-fetching.

  github-contents-pp-cli verify ./books mjwoon/AI-readings/books --agent

• sync-dir — Update an existing downloaded directory in place, fetching only files that changed upstream or are missing locally.

  The weekly-mirror command: keeps a local collection current for the price of one API call plus only the changed bytes.

  github-contents-pp-cli sync-dir ./books mjwoon/AI-readings/books --agent

Remote inspection

• stats — Size and file-count breakdown of any repo path by subfolder and extension, plus the largest files, from a single API request.

  Answers 'how big is this directory and what is in it' in one call before committing to a download.

  github-contents-pp-cli stats mjwoon/AI-readings/books --agent --select by_folder

• search — Search previously fetched repo listings from the local store — find files by name or pattern with zero API calls.

  Lets agents answer 'which repo path had file X' from data already on disk instead of re-walking the API.

  github-contents-pp-cli search "transformers" --limit 20

Command Reference

blobs — Raw git blobs — fetch file content by SHA (works for files up to 100 MB)

• github-contents-pp-cli blobs <owner> <repo> <file_sha> — Get a blob (base64-encoded content) by its git SHA

contents — List directories and read files via the repository contents API

• github-contents-pp-cli contents <owner> <repo> <path> — Get a directory listing (JSON array) or file metadata+content (JSON object, base64, files up to 1 MB).

rate-limit — API quota status — check headroom before and during bulk downloads

• github-contents-pp-cli rate-limit — Show remaining API quota (5000/h with a token, 60/h without)

releases — Repository releases and their downloadable assets

• github-contents-pp-cli releases latest — Get the latest published release
• github-contents-pp-cli releases list — List releases with their assets
• github-contents-pp-cli releases download <owner> <repo> — Download release assets (latest, or --tag) matching --pattern into --out; --list-only previews matches without downloading

repos — Repository metadata (default branch, visibility, size)

• github-contents-pp-cli repos branches — List branches of a repository
• github-contents-pp-cli repos commits — List commits, optionally filtered to a path (newest first)
• github-contents-pp-cli repos get — Get repository metadata including the default branch

tarball — Full-repository snapshot as a .tar.gz

• github-contents-pp-cli tarball <owner/repo[#ref]> — Download the whole repo at a ref as a tarball (use fetch for a subdirectory)

trees — Git trees — one-request recursive listings of an entire repo or subtree

• github-contents-pp-cli trees <owner> <repo> <tree_sha> — Get a git tree. With --recursive, returns every file under it in one request (check the truncated flag on huge repos)

Finding the right command

When you know what you want to do but not which command does it, ask the CLI directly:

github-contents-pp-cli which "<capability in your own words>"

which resolves a natural-language capability query to the best matching command from this CLI's curated feature index. Exit code 0 means at least one match; exit code 2 means no confident match — fall back to --help or use a narrower query.

Recipes

Download a folder, keep the structure

github-contents-pp-cli fetch mjwoon/AI-readings/books --out ./books

Recursive download of one subdirectory; subfolders and filenames (including spaces) arrive intact.

Preview cost before a big download

github-contents-pp-cli plan mjwoon/AI-readings/books --agent

Machine-readable plan: files, total_bytes, api_cost, lfs_pointers — decide before spending bandwidth.

Only PDFs, nothing else

github-contents-pp-cli fetch mjwoon/AI-readings/books --include "*.pdf" --out ./pdfs

Glob filters restrict the download without changing the preserved structure.

Narrow a huge tree listing for an agent

github-contents-pp-cli trees mjwoon AI-readings main --recursive --agent --select tree.path,tree.size

The recursive tree of a repo is thousands of entries; --select keeps only the two fields an agent needs.

Keep a mirror current

github-contents-pp-cli sync-dir ./books mjwoon/AI-readings/books --agent

One API call to diff by blob SHA, then only changed or new files are streamed down.

Auth Setup

Auth is optional: public repos work unauthenticated (60 req/h); a token raises the limit to 5000 req/h and enables private repos. The persistence path is the environment variable:

export GITHUB_TOKEN="your-token-here"   # or GH_TOKEN

github-contents-pp-cli auth setup prints these generic env-var instructions (this CLI has no configured token-creation URL, so --launch has nothing to open). Verify with github-contents-pp-cli auth status or github-contents-pp-cli doctor.

Agent Mode

Add --agent to any command. Expands to: --json --compact --no-input --no-color --yes.

• Pipeable — JSON on stdout, errors on stderr

• Filterable — --select keeps a subset of fields. Dotted paths descend into nested structures; arrays traverse element-wise. Critical for keeping context small on verbose APIs:

  github-contents-pp-cli blobs mock-value mock-value mock-value --agent --select id,name,status

• Previewable — --dry-run shows the request without sending

• Offline-friendly — sync/search commands can use the local SQLite store when available

• Non-interactive — never prompts, every input is a flag

• Read-only — do not use this CLI for create, update, delete, publish, comment, upvote, invite, order, send, or other mutating requests

Response envelope

Commands that read from the local store or the API wrap output in a provenance envelope:

{
  "meta": {"source": "live" | "local", "synced_at": "...", "reason": "..."},
  "results": <data>
}

Parse .results for data and .meta.source to know whether it's live or local. A human-readable N results (live) summary is printed to stderr only when stdout is a terminal AND no machine-format flag (--json, --csv, --compact, --quiet, --plain, --select) is set — piped/agent consumers and explicit-format runs get pure JSON on stdout.

Paths and state

Agents should treat the CLI's path resolver as part of the runtime contract:

• Use --home <dir> for one invocation, or set GITHUB_CONTENTS_HOME=<dir> to relocate all four path kinds under one root.

• Use per-kind env vars only when a specific kind must diverge: GITHUB_CONTENTS_CONFIG_DIR, GITHUB_CONTENTS_DATA_DIR, GITHUB_CONTENTS_STATE_DIR, GITHUB_CONTENTS_CACHE_DIR.

• Resolution order is per-kind env var, --home, GITHUB_CONTENTS_HOME, XDG (XDG_CONFIG_HOME, XDG_DATA_HOME, XDG_STATE_HOME, XDG_CACHE_HOME), then platform defaults.

• config contains settings like config.toml and profiles. data contains credentials.toml, data.db, cookies, and auth sidecars. state contains persisted queries, jobs, and teach.log. cache contains regenerable HTTP/cache files.

• Stored secrets live in credentials.toml under the data dir. Existing legacy config.toml secrets are read for compatibility and leave config.toml on the first auth write.

• Run github-contents-pp-cli doctor --fail-on warn to surface path and credential-location warnings. agent-context exposes a schema v4 paths block for agents that need the resolved dirs.

• For MCP, pass relocation through the MCP host config. The MCP binary does not inherit CLI flags:

  {
    "mcpServers": {
      "github-contents": {
        "command": "github-contents-pp-mcp",
        "env": {
          "GITHUB_CONTENTS_HOME": "/srv/github-contents"
        }
      }
    }
  }

Fleet precedence: an inherited per-kind env var overrides an explicit --home for that kind. Use GITHUB_CONTENTS_HOME or per-kind vars as durable fleet levers, and use --home only for a single invocation. Relocation is not reversible by unsetting env vars; move files manually before clearing GITHUB_CONTENTS_HOME, or doctor will not find credentials left under the former root.

Automatic learning

This CLI ships a self-capturing learning loop. The CLI does its own bookkeeping: every invocation is journaled locally, a failed flag followed by a corrected retry auto-derives a flag_alias candidate, and a teach on a query family without a playbook auto-synthesizes a playbook_candidate from the session's journal. Your job is judgment only: recall first, act on surfaced candidates, teach the final answer, playbook amend when you observe a correction. You never record failures by hand.

Step 1: recall before any discovery

Before list/search/drill commands on a new user question, run:

github-contents-pp-cli recall "<user's question>" --agent

The response envelope:

{
  "query": "...",
  "normalized": "<normalized form>",
  "query_entities": ["..."],
  "found": true | false,
  "match_score": 0.0,
  "results": [
    { "resource_id": "...", "resource_type": "...", "venue": "...",
      "confidence": 2, "entity_match": "exact|partial|unknown",
      "source": "taught|preseed|pattern", "warnings": ["..."] }
  ],
  "mismatches": [ /* only when --debug-mismatches */ ],
  "warnings": [ /* top-level */ ],
  "candidates": [
    { "id": 12, "class": "flag_alias | playbook_candidate",
      "summary": "...", "sightings": 3, "last_seen": "...",
      "rationale": "...",
      "next_action": ["<trial command>", "github-contents-pp-cli learnings confirm 12"] }
  ],
  "playbook": {
    "query_family": "...",
    "playbook": {
      "steps": [ { "cmd": "<command with {slot} substitution>", "purpose": "..." } ],
      "entity_slots": ["$ENTITY"],
      "expected_tool_calls": 3
    },
    "slots_resolved": { "$ENTITY": { "token": "<live token>", "canonical": "<canonical>" } },
    "notes": "<workarounds + gotchas for this query family>"
  },
  "notes": "<duplicate surface for non-playbook callers>"
}

Empty-store short-circuit: if the store has no learnings, playbooks, or candidates yet (recall finds nothing and learnings list and learnings candidates are both empty), skip recall for the rest of this session instead of taxing every query; resume recall-first once something has been taught.

Step 2: decision tree

Read candidates, playbook, notes, results[0], and warnings in that order:

if Candidates present (warnings include "candidates_present"):
    -> candidates are try-then-confirm, never facts. Follow each candidate's
       two-step next_action verbatim: run the trial command first, then run
       `learnings confirm <id>` only after the trial verified the behavior.
       Reject a wrong candidate with `learnings reject <id>`.
    -> NEVER re-teach something recall surfaced as a candidate; confirm or
       reject that candidate instead of teaching a duplicate.
    -> candidates ride alongside playbooks and resource hits, not instead of
       them; continue with the branches below after acting on them.

if Playbook present:
    -> READ Playbook.notes verbatim FIRST (workarounds + gotchas the CLI surface doesn't expose)
    -> replay Playbook.steps in order, substituting Playbook.slots_resolved entries
       for the entity slot tokens. If a step's slot is unresolved, fall back to
       discovery for that step only.
    -> the Playbook's expected_tool_calls is a budget; if you find yourself running
       materially more, record the divergence via `github-contents-pp-cli playbook amend`
       at end-of-session.

elif Notes present (no Playbook):
    -> read Notes verbatim before any discovery step; they carry known gotchas
       for this query family even when no structured choreography exists yet.

elif Found AND Results[0].EntityMatch == "exact" AND Results[0].Confidence >= 2:
    -> skip discovery; fetch live data for Results[*].ResourceID in parallel

elif Found AND Results[0].EntityMatch == "partial":
    -> candidate hint, NOT a hit; read the resource title to validate before trusting

elif (any row in Mismatches[] when --debug-mismatches was passed):
    -> treat as cold start; the stored learning is for a different entity
       (different canonical resolved from query_entities)

else:  // Found == false, no playbook, no notes
    -> cold start; run discovery normally; teach the answer afterward (Step 4).
       If the family has no playbook yet, that teach auto-synthesizes a
       playbook candidate from this session's journal - you do not need to
       record one by hand.

Playbook and Notes are orthogonal to the per-resource path. A recall response can carry both a Playbook AND a Results[] hit - use both: the Playbook tells you which choreography to run; the resource hits short-circuit specific steps. Default to skipping mismatches; pass --debug-mismatches only when investigating cold-start surprises.

Candidate judgment details: learnings confirm <id> prints the candidate's full payload before materializing it - check that the printed payload matches the behavior you verified. learnings reject <id> tombstones the derivation signature so the same candidate does not resurface. The envelope carries only the few candidates worth acting on now; github-contents-pp-cli learnings candidates lists the full open set.

Graceful degradation: if learnings confirm is an unknown command, you are driving an older binary - ignore the candidates guidance and follow the rest of the protocol.

Step 3: always read warnings

• low_confidence: row exists at confidence<2. Treat as a hint, not a skip-discovery hit.
• resource_not_in_store: the local store doesn't have the resource the learning points at. The match validator couldn't classify entities — direct-fetch and re-evaluate.
• cross_alias_match (per-result): the row was taught under a different alias and matched the live query's canonical via entity_lookups (e.g., a "USA" teach satisfying a "United States" recall). Trust the resource_id.
• similar_shape_different_entity:<canonical> (top-level): a structurally matching row exists but its canonical entity differs from the live query's. Treated as cold start; the warning carries the conflicting canonical as a hint, but the row is NOT promoted into Results.
• ambiguous_alias (top-level): a single query entity resolved to multiple canonicals (e.g., "Cards" → Arizona Cardinals + St. Louis Cardinals). Surface the ambiguity from context before committing to a resource.
• candidates_present (top-level): the envelope carries a candidates section. Handle it via the candidates branch in Step 2 before anything else.
• lookup_refresh_available (top-level): an entity in the query has no lookup row yet, but locally stored data could provide one. Re-run plan, fetch, or sync-dir against the target to refresh the local store (this CLI has no standalone sync command).
• Top-level no_learnings_for_query_family: the table had no rows above the Jaccard floor. Pure cold start.