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 tjcoder-labs/cli --skill android-debloatgit clone --depth 1 https://github.com/tjcoder-labs/cliWrote 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/tjcoder-labs/cli/android-debloat)<a href="https://agentmods.dev/skills/tjcoder-labs/cli/android-debloat"><img src="https://agentmods.dev/badge/skills/tjcoder-labs/cli/android-debloat/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/tjcoder-labs/cli/android-debloat"><img src="https://agentmods.dev/badge/skills/tjcoder-labs/cli/android-debloat.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 | $0.00000 | $0.01310 |
| Opus 5.5 | $0.00000 | $0.00524 |
| Sonnet 5.5 | $0.00000 | $0.00262 |
| Haiku 4.5 | $0.00000 | $0.00131 |
Grade A, and why
android-debloat 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 — 136 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Proposal: stored, reusable agent skills
Summary
Add skills, not an Android-specific debloat tool. A skill is a versioned
Markdown playbook with a small metadata header, discovered by the CLI and loaded
into an agent only when relevant. The first skill is Android debloat; its
operational knowledge lives in the debloat workspace's DEBLOAT.md.
The CLI already supports stored workspace agents in
.ergo-cli-go/agents.json and auto-injects workspace AGENTS.md/CODER.md.
Those mechanisms are useful precedents but are not skill support: an agent is a
persona/tool selection, and repository instructions are always loaded and capped
at 12,000 characters. Do not put the whole device-specific playbook in the global
Android assistant prompt.
Skill format and storage
Use a conventional directory with one required Markdown file per skill:
~/.config/tjcoder/coder-cli/skills/
android-debloat/
SKILL.md
references/
motorola-nevada-tracfone.md
Workspace-local skills can live under .ergo-cli-go/skills/ and override or
supplement user skills by name. The global location follows appdir.Dir() rather
than introducing another configuration root. Initially ship documentation and a
loader; optionally ship android-debloat as a built-in embedded skill.
SKILL.md starts with simple YAML front matter:
---
name: android-debloat
description: Audit and plan conservative ADB debloating; use for bloatware,
preinstalled apps, carrier bundles, package removal, disable, or rollback.
version: 1
---
The body contains only portable procedure, decision criteria, safety rules, and
links to references. Device inventories and personal observations are separate
reference documents, not automatically injected on every Android conversation.
DEBLOAT.md can be split into SKILL.md plus device references, or copied into
the skill with attribution and a clear source/version. Prefer one maintained
source over silently diverging copies.
Keep the format deliberately small: Markdown body, front matter, optional references. No scripts, executable plugins, package manager, remote fetching, automatic shell evaluation, or Android-specific Go tool. A skill never grants new tools or permission by itself.
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 · 136 lines · 0 tokens per session scan A 0ba90383cce9
android-debloat is a skill published in the GitHub repository tjcoder-labs/cli (13 stars, last pushed 2d ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 1,310 tokens. 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 2026-11-14.
android-debloat compared with skills from other repositories
| Mod | Type | Stars | Last push | Licence | Scan | Tokens | Works with |
|---|---|---|---|---|---|---|---|
| android-debloattjcoder-labs/cli | Skill | 13 | 2d ago | MIT | A | 0 | Claude Code, Codex |
| imagegenopenai/codex | Skill | 135k | today | Apache-2.0 | A | 113 | Codex |
| playwright-climicrosoft/playwright | Skill | 100k | today | Apache-2.0 | A | 19 | Claude Code |
| browser-automationopenclaw/openclaw | Skill | 395k | today | MIT | A | 36 | Claude Code |
| local-ai-agentsmicrosoft/ai-agents-for-beginners | Skill | 80k | 8d ago | MIT | A | 200 | Claude Code, Codex |
| next-cache-components-adoptionvercel/next.js | Skill | 144k | today | MIT | A | 95 | Claude Code |
| chroniclemicrosoft/vscode | Skill | 195k | today | MIT | A | 72 | Claude Code, Codex |