Borrowing it
Nothing to install: this file belongs to jacob-dietle/context-os. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.
curl -O https://raw.githubusercontent.com/jacob-dietle/context-os/main/.claude/skills/code-service-defrag/SKILL.mdgit clone --depth 1 https://github.com/jacob-dietle/context-osWrote 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/jacob-dietle/context-os/code-service-defrag)<a href="https://agentmods.dev/skills/jacob-dietle/context-os/code-service-defrag"><img src="https://agentmods.dev/badge/skills/jacob-dietle/context-os/code-service-defrag.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 5.1 | $0.00163 | $0.03498 |
| Opus 5 | $0.00081 | $0.01749 |
| Sonnet 5 | $0.00033 | $0.00700 |
| Haiku 4.5 | $0.00016 | $0.00350 |
Grade A, and why
code-service-defrag 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 6d 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 — 220 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Code + Service Defrag
Periodic consolidation of a fragmenting codebase — code AND context — back into coherent, single-source-of-truth wholes. The maintenance discipline that counteracts the entropy agentic coding produces.
Meta-Principle: Agentic coding fragments state faster than humans consolidate it. Fragmentation only accumulates — it is the second law applied to repos. Defrag is the periodic counter-force. Run it on a cadence, or discover the fragmentation the day a deploy fires from the wrong directory.
Why Fragmentation Is Inevitable (it is not a failure)
Conway's Law says a system mirrors the communication structure of whoever built it. Invert it: a codebase touched by parallel agents and multiple humans across many sessions — with no single mind holding the whole map — necessarily accumulates fragments. Duplicate configs, forks, orphan specs, services quietly sharing a database. This is not sloppiness to feel bad about; it is the expected entropy of high-leverage, multi-actor development. The only real question is whether you defrag periodically or meet the fragmentation when it detonates.
The leverage of agentic coding (ship 10x faster) is paid for in visibility (nobody holds the full map). Defrag buys the visibility back, cheaply, on a schedule.
Two Terrains Fragment in Parallel
| Terrain | Fragments into | Detonates as |
|---|---|---|
| Code | duplicate deploy targets, colliding bindings, stale forks, orphan repos | a silent prod overwrite — deploy from the wrong dir clobbers the live one |
| Context | parallel spec conventions, orphan docs, scattered context packages | a false foundation — an agent onboards from the wrong context and builds on it |
Treat both as first-class. Code drift fails loud (eventually — when a deploy detonates). Context drift fails silent — an agent reads the wrong spec, produces confidently wrong work, and nobody notices until it ships. The context half of this skill is not a footnote to the code half; it is the half that fails without a stack trace.
What ships with it
11 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.
- references/cadence.md 3.2 KB
- references/downstream-handoff.md 4.9 KB
- references/landmine-patterns.md 8.1 KB
- references/platform-collision-semantics.md 4.9 KB
- scripts/defrag_scan.sh 1.6 KB runs code
- scripts/scan_archive_markers.sh 1.6 KB runs code
- scripts/scan_bindings.sh 4.4 KB runs code
- scripts/scan_deploy_surface.sh 3.8 KB runs code
- scripts/scan_deploy_target_names.sh 6.2 KB runs code
- scripts/scan_spec_dirs.sh 4.9 KB runs code
- scripts/scan_stale_forks.sh 2.7 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.
- 6d ago First seen · 220 lines · 163 tokens per session scan A f0b234c25a00
code-service-defrag is a skill published in the GitHub repository jacob-dietle/context-os (108 stars, last pushed 23d ago), licensed MIT. It adds 163 tokens to every session and 3,498 once invoked, about $0.0008 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
debug-optimize-lcp
Guides debugging and optimizing Largest Contentful Paint (LCP) using Chrome DevTools MCP tools. Use this skill whenever the user asks about LCP performance, slow page loads, Core Web Vitals optimization, or wants to understand why their page's main content takes too long to appear. Also use when the user mentions…
systematic-debugging
Use when debugging a failing test, build error, or runtime issue that isn't immediately obvious. Guides a 4-phase root cause analysis instead of random fix attempts.
diagnose
Trace from a reproduced symptom to the source code that causes it. Pin the specific file and approximate line, rate confidence in the cause and clarity of the fix independently, and always propose a concrete fix.
repro-admin
Reproduce an EmDash admin UI bug. Attach a container, start the demo dev server, drive the admin with agent-browser using the dev-bypass session, and capture the reproduction as screenshots plus a replayable transcript.
log-error-digest
Analyze log files to troubleshoot errors, identify peak error periods, and produce error clustering, frequency statistics, and time distribution reports. Supports JSON, syslog, and Nginx formats with automatic detection. Use when a user uploads a .log file and asks to analyze errors, find patterns, debug issues, or…
byted-util-volcengine-detect-retry
Orchestrates Volcengine Cloud Detect (云拨测) false-alarm reduction for failing periodic-task dial points. Use for scheduled task scans, retrying failed nodes, control-testing node health, distinguishing real target outages from broken dial points, and suggesting same-type LastMile/IDC/Private replacement nodes. Runs the…