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/joshukraine/dotfiles/sidecarnpx skills add joshukraine/dotfiles --skill sidecargit clone --depth 1 https://github.com/joshukraine/dotfilesWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/joshukraine/dotfiles/sidecar)<a href="https://agentmods.dev/skills/joshukraine/dotfiles/sidecar"><img src="https://agentmods.dev/badge/skills/joshukraine/dotfiles/sidecar.svg" alt="Measured on agentmods" height="20"></a>What 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 | $0.00074 | $0.00829 |
| Opus 5 | $0.00037 | $0.00415 |
| Sonnet 5 | $0.00015 | $0.00166 |
| Haiku 4.5 | $0.00007 | $0.00083 |
Grade A, and why
sidecar 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 5d 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.
How it starts
The opening of the file, as written. The whole thing — 42 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Sidecar (research-only session)
You are a secondary session sharing a working directory with another Claude Code session that is actively editing code in the same repository. Your job here is research, planning, theory, discussion, and drafting — not implementation. The other session owns the working tree; you must not step on it.
Read-only operations, web/MCP research, and outward-facing actions that touch no local state are all fine. The only things that cause collisions are mutations to shared local state, so those are off-limits.
The one rule
Do not mutate shared local state. That state is exactly: tracked files in the working tree, and git's index/HEAD. Everything dangerous follows from those two; everything safe is safe because it touches neither.
Forbidden
- Editing, creating, or deleting tracked files in the repository (no
Edit,Write, orNotebookEditon anything under version control). Code, config, docs — all off-limits. - Git state mutations:
git add,git commit,git checkout/git switch,git branch,git stash,git reset,git restore,git rebase,git merge,git pull. These corrupt whatever the other session is doing with the index, working tree, or HEAD. - Creating branches or pull requests (they require branch + push, i.e. git mutation).
- Installing dependencies, running formatters that rewrite files, or starting long-running servers — these mutate
node_modules, lockfiles, the tree, or fixed ports the other session may rely on.
Allowed
- Read anything — files, code, docs, history.
- Read-only git:
git log,git diff,git status,git show,git blame. - Web search / fetch, Context7, and MCP tools — research freely.
- Write to the session scratchpad — your scratchpad directory is session-isolated (its own UUID'd path), so temp files never collide with the other session. Put all temp files there, never in the repo root (e.g. do not write
.git_commit_msgto the repo root; that path is shared). - Create and comment on GitHub issues via
gh— this is outward-facing but touches no local shared state. When an issue body needs a temp file, write it to the scratchpad, thengh issue create -F <scratchpad-path>.
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.
- 5d ago First seen · 42 lines · 74 tokens per session scan A 52f02a24a9c1
sidecar is a skill published in the GitHub repository joshukraine/dotfiles (422 stars, last pushed 23d ago), licensed MIT. It adds 74 tokens to every session and 829 once invoked, about $0.0004 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
cmux-rebuild
Manage the user's durable dev sessions — zellij sessions on remote hosts (bonbon, taffy) reached over mosh, plus the local host's (trifle) own detached zellij sessions, surfaced as cmux tabs via ssh::durable / zellij::resume. Load when the user wants to rebuild/resurrect lost cmux durable surfaces after a cmux restart…
ml-paper-writing
Write publication-ready ML/AI papers for NeurIPS, ICML, ICLR, ACL, AAAI, COLM. Use when drafting papers from research repos, structuring arguments, verifying citations, or preparing camera-ready submissions. Includes LaTeX templates, reviewer guidelines, and citation verification workflows.
python-conventions
Python coding conventions, modern tooling, and project setup. Apply when writing or reviewing Python code, creating projects, writing scripts, or migrating from legacy tools. Covers uv, ruff, ty, type hints, code structure, async, testing, and project configuration.
review-pr
Review an existing GitHub PR — one verified single-pass read of the diff, inline P1–P3 findings, fixes for the P1/P2s pushed to the PR branch, scoped verification, resolved threads, a summary. Escalates to a two-reader second opinion only for large source changes. Use when asked to review and fix a PR, or to run the…
slidesync
Bidirectional sync between a Slidev markdown deck and Google Slides as native, editable objects (not images). Push markdown -> Slides, pull Slides -> markdown, and round-trip. Use when the user wants to author/version a Google Slides deck from markdown, replicate a deck's styling, or keep a .slidev.md in sync with a…
cmux-fork-session
Fork the current agent session — Claude Code or pi — into a new cmux split pane (or tab): opens a split beside the caller, relaunches this session forked (claude --fork-session / pi --fork), titles it 'fork: ', and keeps that title. Use when the user asks to fork/duplicate/branch the current session into a new split…