agentmods.dev

ctx-core

ctx-core is a skill for Claude Code from broomva/skills. It costs 374 tokens per session (7,497 once invoked), scanned B, original, MIT.

A read-only shared board for Claude Code sessions in one workspace scope. It records structured session status, location, branch, and whether a session stopped or died, then shows relevant information when a session starts.

In plain words
What is it for?
It is for tracking which agent sessions exist, where they are working, their latest status keyword, and whether they ended unexpectedly.
Why use it?
It keeps basic coordination state on disk so it is less likely to disappear from an agent's conversation history and can be seen by other sessions.

Skill for Claude Code

Written for Claude Code: when-to-use in frontmatter. Also seen: reads .claude/ paths; mentions subagents; mentions Claude Code.

Not installable: its command points at a path on the author’s own machine, so it runs nowhere else. The line is /Users/broomva/broomva/skills/skills/orchestration/ctx-core/scripts/ctx-hook.sh.

Part of the orchestration plugin — 14 skills shipped together , and of broomva-skills

not rated 5repo changed 23d ago B scan 374 tokens original MIT

Good fit It is for tracking which agent sessions exist, where they are working, their latest status keyword, and whether they ended unexpectedly.

Compare 6 skills from other repositories ↓
View source ↗ broomva/skills skills.sh
Install

Getting it into your agent

This one installs as part of its plugin. Adding the marketplace and installing the plugin brings it with everything else the plugin ships.

Claude Code
/plugin marketplace add broomva/skills
Claude Code
/plugin install orchestration

Made for: Claude Code.

Or install orchestration, the plugin that ships this one along with the rest of its 14 skills.

Wrote 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.

agentmods badge for ctx-core

README.md
[![agentmods](https://agentmods.dev/badge/skills/broomva/skills/ctx-core/github.svg)](https://agentmods.dev/skills/broomva/skills/ctx-core)
Your own site
<a href="https://agentmods.dev/skills/broomva/skills/ctx-core"><img src="https://agentmods.dev/badge/skills/broomva/skills/ctx-core/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.

agentmods 80×15 button for ctx-core

Your own site · 80×15
<a href="https://agentmods.dev/skills/broomva/skills/ctx-core"><img src="https://agentmods.dev/badge/skills/broomva/skills/ctx-core.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 374 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 7,497 The whole file, excluding the scripts and references it only reads on demand.
Security scan B 2 findings. A grade says what 26 rules found in the file — not that it is safe.
Origin original No closer match found in the catalogue.
Token cost

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.

ModelPer sessionOnce invoked
Fable 5.1 $0.00374 $0.07497
Opus 5.5 $0.00150 $0.02999
Sonnet 5.5 $0.00075 $0.01499
Haiku 4.5 $0.00037 $0.00750

Measured 23d ago against content hash 133ea170b3ac, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade B, and why

ctx-core scanned grade B with 2 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 23d ago.

The scan reads SKILL.md. This mod also ships 36 executable files (scripts/ctx_compare.py, scripts/ctx_hook.py, scripts/ctx_keys.py, …), listed below but not scanned — reading those needs a real analyzer, not pattern matching.

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.

Reads agent configuration directoriesmediumAgent snooping

.claude/, .codex/, .gemini/ hold keys, settings and other credentials a mod has no legitimate need for.

Registering hooks is human-gated. No agent edits `~/.claude/settings.json`,

Makes network callslowCapability

Not a fault in itself. Listed so you know the mod talks to something, and to what.

- `curl` the Paseo MCP endpoint with the bearer token in its own argv. The
skills/orchestration/ctx-core/SKILL.md · 443 lines

How it starts

The opening of the file, as written. The whole thing — 443 lines — stays where its author put it; the contents beside it link to each section on GitHub.

ctx-core: the shared board (phase 1)

In the 09-26 → 09-29 run, coordination state lived in the coordinator's context window. Compaction kept dropping it, and no worker session could see another. Phase 1 of the shared context core (design, docs/specs/2026-09-29-shared-context-core.html) puts the minimum of that state on disk: which sessions exist in a scope, where they are (cwd, branch), the keyword of their last ARC-STATUS line, and whether they died. Every session gets the relevant slice of it at start.

It is read and publish only. There is no mailbox, no wake-up, and no retention.

What it stores: structured fields, no free text

Field From Kept as
session_id, paseo_agent_id hook input, PASEO_AGENT_ID id-shaped strings ([A-Za-z0-9][A-Za-z0-9._:-]{0,127}), otherwise omitted
cwd, repo, branch the filesystem and git flattened to one line
type, ts the hook session.start, session.stop or session.died; UTC milliseconds
arc_status, arc_line Stop's last_assistant_message The last line matching exactly ^ARC-STATUS: [A-Z]+. The keyword is MERGED, CLOSED, DONE, BLOCKED, or OTHER, plus at most the first 120 characters of that line
error (the row's died_error) StopFailure The error class (rate_limit, billing_error, …). Claude Code 2.1.280 sends it as error; the reference documents it as error_type. A value that is not class-shaped is stored as unknown. error_details and the message text are never read

Nothing else is written: no model, source, role, prose, message text or error text. The schema is enforced on read too, so a hand-written log line with free text, an unknown key, or a guard-failing field is skipped by the fold. Schema: references/event-schema.md.

One cheap guard (guard_ok) runs on every stored string. It drops a field in either of two cases:

  • The field contains crm/, Bearer , ghp_, gho_, github_pat_, sk-, xox or AKIA. These match case-insensitively, and only at the start of a token, so fix/task-list and mycrm/ pass.
  • The field contains a run of 32 or more ASCII letters and digits.

Read the full file on GitHub · 443 lines

Files

What ships with it

44 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.

Changes

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.

  1. 23d ago Changed 133ea170b3ac
  2. 1mo ago Changed · +106 lines · +107 tokens per session 62400a38c51a
  3. 1mo ago First seen · 337 lines · 267 tokens per session scan B 5971bc59410e

Subscribe to this mod's changes

ctx-core is a skill published in the GitHub repository broomva/skills (5 stars, last pushed 2d ago), licensed MIT. It adds 374 tokens to every session and 7,497 once invoked, about $0.0015 per session on Opus 5.5. A static security scan graded it B with 2 findings (reads agent configuration directories, makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-09-30.

Building agentsClaude CodeMemory and contextOrchestrationagent-skillsai-agentsmcpremotionskills
Related

ctx-core compared with skills from other repositories

ctx-core next to 6 skills from other repositories
ModTypeStarsLast pushLicenceScanTokensWorks with
ctx-corebroomva/skillsSkill52d agoMITB374Claude Code
gossipcatgossipcat-ai/gossipcat-aiSkill4127d agoMITA50Claude Code, Codex
swarmclawswarmclawai/swarmclawSkill7064mo agoMITB67Claude Code, Codex
using-aoUntrivial-ai/agent-orchestratorSkill14k2d agoApache-2.0A71Claude Code, Codex
comet-memoryrpamis/cometSkill3.3ktodayMITA38Claude Code, Codex
relevance-coarse-filterelvisun/newsjackSkill1.7k2d agoMITA57Claude Code
award-calendarborski/travel-hacking-toolkitSkill926todayMITA52Claude Code