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 skills add Ruso-0/Nreki --skill templatesgit clone --depth 1 https://github.com/Ruso-0/NrekiWrote 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/ruso-0/nreki/templates)<a href="https://agentmods.dev/skills/ruso-0/nreki/templates"><img src="https://agentmods.dev/badge/skills/ruso-0/nreki/templates/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/ruso-0/nreki/templates"><img src="https://agentmods.dev/badge/skills/ruso-0/nreki/templates.svg" alt="Reviewed on agentmods" width="80" 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 5.1 | $0.00043 | $0.01268 |
| Opus 5 | $0.00022 | $0.00634 |
| Sonnet 5 | $0.00009 | $0.00254 |
| Haiku 4.5 | $0.00004 | $0.00127 |
Grade A, and why
nreki-optimizer 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 12d 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.
This is a copy
86% identical to nreki-optimizer — 12 lines differ, which has more behind it and is treated as the original. This page carries a canonical link to it rather than competing with it.
How it starts
The opening of the file, as written. The whole thing — 59 lines — stays where its author put it; the contents beside it link to each section on GitHub.
NREKI ACTIVE (IMMUNE SYSTEM)
If nreki_* tools fail: fall back to native Read/Write ONLY for new files or <100L. Native tools are BLOCKED by filesystem hooks for >100L files.
0. ZERO-CHATTER
- Call tools immediately. No "I will now..." or "Let me...".
- State changes in one line. No ASCII tables, no summaries.
- Output tokens cost 5x input. Optimize every syllable.
1. NAVIGATION & READING
- NEVER use Explore, Bash grep/rg/cat, or native Read File.
- EXACT SEARCH: ALWAYS use nreki_navigate action:"fast_grep". Returns 3D AST topological coordinates (File + Symbol + Line) via SQLite index. Benchmarked 5-33x faster than bash grep on indexed codebases.
- SEMANTIC SEARCH: Use nreki_navigate action:"search" for concept-based queries (Type Ledger).
- BATCH READING: NEVER read functions one by one. Use ONE call with comma-separated targets: nreki_code action:"compress" focus:"func1, func2, func3".
- LARGE FILES: nreki_navigate action:"outline" auto-expands HIGH-risk methods up to 6,000 token budget.
- CRITICAL: If outline says [BUDGET LIMIT REACHED], you MUST use compress focus:"<omitted_symbols>" to read the hidden logic before editing. Do not guess.
- Pay attention to [Defect] flags and [orphan / 0 ext refs] markers.
- Small files: nreki_code action:"read".
2. SURGICAL EDITING & ACID BATCHES
- NEVER use native Write/Replace on existing files (bypasses AST validation).
- NEVER edit sequentially. Read ALL targets first, then batch ALL fixes in ONE nreki_code action:"batch_edit" call.
- 80L PAYLOAD GUILLOTINE: NREKI mathematically rejects any single payload (new_code or replace_text) >80 lines. You MUST decompose large rewrites.
- PATCH MODE (MANDATORY): If the original symbol is larger than
SYMBOL_REPLACE_LIMIT(default 100 lines, empirically calibrated in v11.3.1; override viaNREKI_SYMBOL_LIMITenv), mode:"replace" is BLOCKED. You MUST use mode:"patch" with search_text and replace_text.- Tolerant Match: search_text indentation is fuzzy. Focus on unique logical content.
- Multi-Patching: To bypass the 80L limit, send MULTIPLE patch edits to the SAME symbol sequentially in one batch_edit.
- ACID Rule: When multi-patching a symbol, EVERY search_text must match the ORIGINAL unaltered source code. Do not target code injected by previous patches.
- ANTI-SWEEP SHIELD: NEVER use // eslint-disable, @ts-ignore, or # noqa. NREKI will block your edit. Fix the structural flaw.
- RENAMING: Run nreki_navigate action:"prepare_refactor" first.
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.
- 12d ago First seen · 59 lines · 43 tokens per session scan A 93d944f3c980
nreki-optimizer is a skill published in the GitHub repository Ruso-0/Nreki (11 stars, last pushed 3mo ago), licensed Apache-2.0. It adds 43 tokens to every session and 1,268 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. It is 86% identical to nreki-optimizer, differing in 12 lines, and is treated as a copy.
Other skills, from other repositories
agt-policy-authoring
Create and validate a minimal AGT Copilot CLI policy tailored to the repository being inspected.
lsp-local-symbols
Fast file-scoped symbol analysis — find all usages of a symbol within the current file, list all symbols defined in the file, and get type info at a position. Use when you need local-scope analysis without a workspace-wide search.
lsp-edit-symbol
Edit a named symbol without knowing its file or position. Use when you want to change a function, type, or variable by name and don't have exact coordinates. Resolves the symbol to its definition, retrieves its full range, and applies the edit.
lsp-implement
Find all concrete implementations of an interface or abstract type. Use when you need to know what types satisfy an interface, or what subtypes exist before changing a base type.
hugging-face-dataset-creator
Create and manage datasets on Hugging Face Hub. Supports initializing repos, defining configs/system prompts, and streaming row updates. Designed to work alongside HF MCP server for comprehensive dataset workflows.
json-transform
Apply a jq-style filter to transform a JSON payload.