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/joaquimscosta/arkhe-claude-pluginsWrote 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/joaquimscosta/arkhe-claude-plugins/context-prime)<a href="https://agentmods.dev/commands/joaquimscosta/arkhe-claude-plugins/context-prime"><img src="https://agentmods.dev/badge/commands/joaquimscosta/arkhe-claude-plugins/context-prime/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/joaquimscosta/arkhe-claude-plugins/context-prime"><img src="https://agentmods.dev/badge/commands/joaquimscosta/arkhe-claude-plugins/context-prime.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 | $0.00065 | $0.01112 |
| Opus 5 | $0.00032 | $0.00556 |
| Sonnet 5 | $0.00013 | $0.00222 |
| Haiku 4.5 | $0.00006 | $0.00111 |
Grade A, and why
context-prime 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 9d 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 — 104 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Context Prime
Usage
/context-prime- Summarize the entire conversation/context-prime <TOPIC>- Focus on specific topics (e.g.,/context-prime the auth refactor and DB migration)
Context
Topic focus (if provided): $ARGUMENTS
If no arguments provided: Summarize the entire conversation.
Instructions
Analyze the current conversation and produce a highly structured context block — a YAML frontmatter header followed by Markdown body — that can be pasted into a new session for immediate situational awareness by another AI assistant.
What to extract
Review the conversation for:
- What is being worked on (project, feature, problem domain)
- Key decisions made and their rationale
- Current state — what's done, what's in progress, what's blocked
- Skills and subagents invoked during the session (for checkpoint resume)
- Immediate next steps or open questions
- A concrete command the next session should run to continue
If a topic focus was provided, narrow the analysis to only the conversation aspects relevant to that topic. Ignore unrelated threads.
Generating the slug
The {slug} is a short kebab-case identifier (2-5 words, ≤40 chars):
- If
$ARGUMENTSis non-empty: slugify it (lowercase, replace whitespace/punctuation with hyphens, drop stop-words if needed) - Otherwise: auto-derive from the dominant topic of the conversation
Examples: auth-refactor-db-migration, context-prime-improvements, fix-jwt-expiry
Capturing git state
Before producing the output, capture the current branch and short commit so the resume session can reproduce state:
git rev-parse --abbrev-ref HEAD
git rev-parse --short HEAD
If not in a git repo, omit branch/commit fields.
Output constraints
Produce the output as a YAML frontmatter block followed by Markdown sections, using exactly the headers below. The output must be optimized for an LLM to quickly parse and understand the situation.
Use bullet points and concise statements rather than long paragraphs.
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.
- 9d ago First seen · 104 lines · 65 tokens per session scan A 5b8d0c336cd1
context-prime is a command published in the GitHub repository joaquimscosta/arkhe-claude-plugins (21 stars, last pushed 26d ago), licensed MIT. It adds 65 tokens to every session and 1,112 once invoked, about $0.0003 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 commands, from other repositories
instinct
Manage instincts (confidence-based rules from session experience).
forge-compact-task
Compact context with dotforge-standardized hint that preserves decisions, files modified, pending TODOs, behaviors disabled, and last commit. Use at 80% context window instead of waiting for auto-compact.
learning
Meta-commands for capturing session insight back into the system.
cap
Shorthand for /forge capture — register a practice or insight in practices/inbox/.
forge-context-status
Report on current context window usage, cache health, and compaction recommendation. Read-only — does not compact.
export-closedloop-learnings
Exports pending ClosedLoop learnings to global location with deduplication.