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.
git clone --depth 1 https://github.com/m-binimran/dev-packWrote 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/commands/m-binimran/dev-pack/error-states)<a href="https://agentmods.dev/commands/m-binimran/dev-pack/error-states"><img src="https://agentmods.dev/badge/commands/m-binimran/dev-pack/error-states/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/commands/m-binimran/dev-pack/error-states"><img src="https://agentmods.dev/badge/commands/m-binimran/dev-pack/error-states.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 | $0.00038 | $0.00091 |
| Opus 5.5 | $0.00015 | $0.00036 |
| Sonnet 5.5 | $0.00008 | $0.00018 |
| Haiku 4.5 | $0.00004 | $0.00009 |
Grade A, and why
error-states 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.
What it actually says
/error-states $ARGUMENTS
Use the error-states skill for this request: $ARGUMENTS
If no details were provided, ask for what the skill needs first, then follow the skill's process exactly.
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 First seen · 10 lines · 38 tokens per session scan A 5e9d903fbfc8
error-states is a command published in the GitHub repository m-binimran/dev-pack (2 stars, last pushed 4mo ago), licensed MIT. It adds 38 tokens to every session and 91 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 2026-08-31.
Other commands, from other repositories
observe-slo
Manage Service Level Objectives (SLO) — define, activate, retire, track violations. Provided by the Observability First methodology.
learn
Capture learnings from a closed spike and optionally promote them to ADR or seed evolution. Provided by the Exploration methodology.
incident
Manage incident records — blameless postmortem capture with state machine. Provided by the Incident Review methodology.
evolve
Evolve the system when /evaluate fails. Runs Wonder → Reflect → Re-seed cycle with fan-out multi-perspective analysis (Contrarian + Simplifier + Researcher). Stops at convergence (ontology similarity ≥ 0.95).
threat-triggers
View or update trigger keywords/paths that auto-require threat models. Provided by the Threat Model (Lite) methodology.
contract
Advance a parallel-change plan from migrate to contract phase, verifying zero callers reference the old signature. Provided by the Parallel Change methodology.