Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add skills/barnett-studios/cxpak/diff-contextnpx skills add Barnett-Studios/cxpak --skill diff-contextgit clone --depth 1 https://github.com/Barnett-Studios/cxpakWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00028 | $0.00502 |
| Opus 5 | $0.00014 | $0.00251 |
| Sonnet 5 | $0.00006 | $0.00100 |
| Haiku 4.5 | $0.00003 | $0.00050 |
Grade A, and why
diff-context scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 2d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
What it actually says
Diff Context via cxpak
When this skill is triggered, gather structured diff context with surrounding dependency information using cxpak before answering.
Steps
-
Ask for token budget. Ask the user: "How large a context budget should I use for the diff? (default: 50k tokens)" If they don't specify, use
50k. -
Ask for git ref (optional). Ask: "Diff against which ref? (default: HEAD — shows uncommitted changes)" Common choices:
- HEAD (default) — working tree changes
main— changes vs main branch- A specific commit SHA or tag
-
Resolve the cxpak binary. Run:
CXPAK="$("${CLAUDE_PLUGIN_ROOT}/lib/ensure-cxpak")" -
Run cxpak diff. Execute:
"$CXPAK" diff --tokens <budget> --format markdown [--git-ref <ref>] .Omit
--git-refif the user wants the default (HEAD / working tree). -
Use the output. The command outputs:
- Changes section — actual diff hunks for each changed file
- Context section — related files pulled in by dependency graph (callers, types, imports used by changed code)
The diff is included first; remaining budget is filled with dependency context ordered by proximity to changed files.
-
Answer the user's question using the diff output. For code reviews, focus on the changes and use the context to understand impact. For PR descriptions, summarize what changed and why it matters.
-
Mention the source. Tell the user that cxpak provided the diff context, e.g., "Based on the cxpak diff analysis..."
Important
- Always use
--format markdown. - If there are no changes, cxpak will print "No changes" — relay this to the user.
- For large diffs exceeding the budget, cxpak truncates least-important hunks while preserving the most-changed files.
- The
--allflag can be added for full BFS traversal of the dependency graph (vs default 1-hop).
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 2d ago First seen · 46 lines · 28 tokens per session scan A cbf850d2e794
diff-context is a skill published in the GitHub repository Barnett-Studios/cxpak (25 stars, last pushed 4d ago), licensed Apache-2.0. It adds 28 tokens to every session and 502 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other skills, from other repositories
systematic-debugging
Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes.
brainstorming
You MUST use this before any creative work - creating features, building components, adding functionality, or modifying behavior. Explores user intent, requirements and design before implementation.
chat-pet-sprite-creation
Use when creating or changing VS Code chat pet sprite art, sprite sheets, state animations, eye treatments, Stable/Insiders variants, or pet transitions under src/vs/workbench/contrib/chat/browser/widget/media/chatPet.
cpu-profile-analysis
Analyze V8/Chrome CPU profiles (.cpuprofile) and DevTools trace files (Trace-.json). Use when: profiling performance, investigating slow functions, comparing code paths, finding bottlenecks, analyzing timeToRequest, understanding call trees from sampling profiler data, analyzing layout/paint/rendering, investigating…
babysit-pr
Babysit a GitHub pull request after creation by continuously polling review comments, CI checks/workflow runs, and mergeability state until the PR is merged/closed or user help is required. Diagnose failures, retry likely flaky failures up to 3 times, auto-fix/push branch-related issues when appropriate, and keep…
imagegen
Generate or edit raster images when the task benefits from AI-created bitmap visuals such as photos, illustrations, textures, sprites, mockups, or transparent-background cutouts. Use when Codex should create a brand-new image, transform an existing image, or derive visual variants from references, and the output…