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 0xDarkMatter/claude-mods --skill adr-opsgit clone --depth 1 https://github.com/0xDarkMatter/claude-modsWrote 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/0xdarkmatter/claude-mods/adr-ops)<a href="https://agentmods.dev/skills/0xdarkmatter/claude-mods/adr-ops"><img src="https://agentmods.dev/badge/skills/0xdarkmatter/claude-mods/adr-ops/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/0xdarkmatter/claude-mods/adr-ops"><img src="https://agentmods.dev/badge/skills/0xdarkmatter/claude-mods/adr-ops.svg" alt="Reviewed on agentmods" width="80" height="20"></a>- NVIDIA SkillSpector pass
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 | $0.00062 | $0.03463 |
| Opus 5.5 | $0.00025 | $0.01385 |
| Sonnet 5.5 | $0.00012 | $0.00693 |
| Haiku 4.5 | $0.00006 | $0.00346 |
Grade A, and why
adr-ops 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 1mo 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 — 307 lines — stays where its author put it; the contents beside it link to each section on GitHub.
ADR Ops
An Architecture Decision Record (ADR) captures one architectural decision: what was decided, why, what was rejected, and what it costs. ADRs are append-only project memory — they exist so a future maintainer touching a subsystem can recover the reasoning behind its shape without archaeology through git history or chat logs.
This skill encapsulates a battle-tested ADR protocol and generalizes it to any repo.
The default location is docs/adr/, but every script takes --dir so a repo can keep
records anywhere (docs/decisions/, architecture/adr/, …).
When to write an ADR
Write one when a change has any of these properties:
- It constrains future options — a boundary, an invariant, a "we will always / never do X" rule that later work must respect.
- Multiple alternatives were seriously evaluated and the choice is not obvious in hindsight.
- The rationale is non-obvious from the code — the code shows what, the ADR preserves why.
Write one decision per ADR. If a change bundles two separable decisions, write two.
When NOT to write one
- A bug fix, refactor, or feature that follows existing architecture without changing it — that is a commit message.
- A reversible, low-stakes choice — that is a code comment.
- A point-in-time event with no forward constraint (a benchmark run, an incident write-up) — that is an audit/log entry, not an ADR.
Rule of thumb: if someone could plausibly undo this next month without re-litigating a trade-off, it is not an ADR.
Naming, location, numbering
- Path:
<adr-dir>/ADR-NNN-slug.md(default<adr-dir>=docs/adr). NNNis zero-padded three digits, assigned sequentially: the next number ishighest existing + 1. Numbers are never reused, never reordered — a superseded ADR keeps its number forever.slugis short kebab-case naming the subject (oauth-only-auth,per-trial-container).- A protocol/how-to file (e.g.
00_*) sorts above the numbered records and is not part of the sequence.
What ships with it
9 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.
- assets/ADR-template.md 1.2 KB
- references/canonical-format.md 4.8 KB
- references/lifecycle-and-supersession.md 3.0 KB
- scripts/adr-index.sh 8.2 KB runs code
- scripts/adr-init.sh 7.6 KB runs code
- scripts/adr-lint.py 22 KB runs code
- scripts/adr-new.sh 12 KB runs code
- scripts/adr-touching.py 17 KB runs code
- tests/run.sh 24 KB runs code
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.
- 1mo ago Changed · +13 lines 3ffae2c26251
- 1mo ago First seen · 294 lines · 62 tokens per session scan A 8d5549ce11b5
adr-ops is a skill published in the GitHub repository 0xDarkMatter/claude-mods (76 stars, last pushed 11d ago), licensed MIT. It adds 62 tokens to every session and 3,463 once invoked, about $0.0002 per session on Opus 5.5. 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
Vizra ADK Memory System
Implement persistent memory, session context, and vector memory (RAG) for AI agents.
nav-compact
Clear conversation context while preserving knowledge via context marker. Use when user says "clear context", "start fresh", "done with this task", or when approaching token limits.
hive-mind-advanced
Queen-led collective-intelligence coordination with consensus mechanisms and persistent shared memory. Use when orchestrating a large swarm of agents under a lead coordinator, building consensus across agents, or maintaining durable cross-agent memory.
creator-productivity
Task management and workspace memory for creative work sessions. Manages a TASKS.md file for project tracking, a two-tier memory system (CREATOR.md + creator-memory/), and a daily sprint structure. Use when organizing creative projects, tracking content tasks, or starting a focused work session.
memory-prune
Manual reviewable prune of stale memory entries. Lists project entries older than 30 days with content preview; Frank decides per-entry [keep | archive | pin]. Keeps MEMORY.md under the auto-load 200-line truncation. Use when MEMORY.md exceeds 180 lines or session-end after a heavy work week.
creator-intelligence
Transform Claude Code into your personal Jarvis - an AI operating system that knows your projects, preferences, and workflows across sessions. Use this skill when setting up personal AI systems, building context-aware assistants, or creating persistent memory across projects.