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/carlos-algms/agentic.nvim/agentic-testingnpx skills add carlos-algms/agentic.nvim --skill agentic-testinggit clone --depth 1 https://github.com/carlos-algms/agentic.nvimWhat 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 | $0.00051 | $0.00819 |
| Opus 5 | $0.00026 | $0.00409 |
| Sonnet 5 | $0.00010 | $0.00164 |
| Haiku 4.5 | $0.00005 | $0.00082 |
Grade A, and why
agentic-testing 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.
How it starts
The opening of the file, as written. The whole thing — 88 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Agentic Testing
Framework
- mini.test with Busted-style emulation.
- Test files live next to source files as
<module>.test.lua. tests/holds the runner, helpers, functional tests, and integration tests.- The previous Busted/lazy.nvim setup is gone.
TDD
For every bug fix or behavioral change:
- Bootstrap missing symbols first so the test loads.
- Write the failing assertion.
- Run the focused test with
make test-file FILE=<path>and confirm it fails for behavior, not setup:- wrong: missing module, nil method, syntax error, unresolved import
- right: value/state/output mismatch
- Implement the minimum code to pass.
- Re-run
make test-file FILE=<path>. - Run the relevant full check.
- After adding or changing tests, reconcile the case count. This guards against
silently dropped OR unexpectedly generated tests. Both are failures.
- Compute EXPECTED cases by reading the file, not grepping:
- each static
it()= 1 case - each
it()inside afor/each/table-driven loop = the loop's iteration count (a singleit(line can emit many cases, or zero) - a grep of
it(is a lower bound, never the answer
- each static
- Read ACTUAL from
Total number of cases: Nin themake test-file FILE=<path>output. - EXPECTED MUST equal ACTUAL. Mismatch = a test was dropped, a loop is empty, or a generator misfired; stop and reconcile before claiming green. Eyeballing ACTUAL alone is NOT the check - you must derive EXPECTED first.
- Compute EXPECTED cases by reading the file, not grepping:
Pure refactors, formatting, and docs can skip red/green, but say that in the PR.
Commands
make test
make test-file FILE=lua/agentic/acp/agent_modes.test.lua
Use make test-file FILE=<path> for the red/green inner loop; it runs one file
in seconds. Never run make validate between iterations. After all .lua
edits, run make validate and fix until it passes - the task is not done until
it does. Pre-commit gate, not a per-test step.
For Lua or test changes, root AGENTS.md requires make validate after the
focused checks.
What ships with it
3 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
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 · 88 lines · 51 tokens per session scan A 5b112144abe0
agentic-testing is a skill published in the GitHub repository carlos-algms/agentic.nvim (609 stars, last pushed 9d ago), licensed MIT. It adds 51 tokens to every session and 819 once invoked, about $0.0003 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
add
Categorize new Neovim plugins and insert them into the correct markdown category files.
write-teach-commit
Draft teach: commit messages for .github/workflows/improve-add-skill.yml. Use when preparing or reviewing a correction commit that should trigger the Improve Add Skill workflow from $add categorization feedback, including moved plugin entries, removed or excluded plugins, changed .reason.md categories, or…
improve-add-skill
Improve the repo-local add skill from reviewed Neovim plugin categorization corrections. Use when a correction commit, PR review result, diff, or artifact such as a teach: commit shows that $add placed, excluded, or reasoned about plugins incorrectly and the reusable lesson should be folded back into…
git-branch-commit-pr
Automates the workflow of creating a feature branch, committing changes, and opening a PR. Use this when asked to create a branch, make changes, commit, and open a pull request for Lua code improvements or fixes.
source-drafting
Use when drafting documents, reports, posts, presentations, or review comments from source materials, or conducting research across provided documents. Enforces read-first drafting, a source map for every claim, agent verification, and separation of verified from unverified claims.
git-commit
Create a new git commit for staged changes. Use when the user asks to commit staged changes with an auto-generated Conventional Commits message, or when Claude Code itself wants to run git commit.