shared-language

shared-language is a skill for Claude Code, Codex from znlgis/my-opencode-deepseek-config. It costs 41 tokens per session (477 once invoked), scanned A, original, MIT.

A project glossary stored in `.opencode/CONTEXT.md`, with one-sentence definitions for terms that have a specific meaning in the codebase.

In plain words
What is it for?
Use it to record project-specific terms and decisions after design discussions, then help agents use those definitions when reasoning about the code.
Why use it?
It prevents coding agents from repeatedly re-explaining the same concepts across messages or sessions. It also gives them a shared reference for project vocabulary.

Skill for Claude CodeCodex

Install

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.

agentmods
npx agentmods add skills/znlgis/my-opencode-deepseek-config/shared-language
Any agent
npx skills add znlgis/my-opencode-deepseek-config --skill shared-language
Clone the repo
git clone --depth 1 https://github.com/znlgis/my-opencode-deepseek-config

Made for: Claude Code, Codex.

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 shared-language

README.md
[![agentmods](https://agentmods.dev/badge/skills/znlgis/my-opencode-deepseek-config/shared-language.svg)](https://agentmods.dev/skills/znlgis/my-opencode-deepseek-config/shared-language)
Your own site
<a href="https://agentmods.dev/skills/znlgis/my-opencode-deepseek-config/shared-language"><img src="https://agentmods.dev/badge/skills/znlgis/my-opencode-deepseek-config/shared-language.svg" alt="Measured on agentmods" height="20"></a>
Per session 41 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 477 The whole file, excluding the scripts and references it only reads on demand.
Security scan A 0 findings. Scan, not verified.
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 $0.00041 $0.00477
Opus 5 $0.00020 $0.00238
Sonnet 5 $0.00008 $0.00095
Haiku 4.5 $0.00004 $0.00048

Measured 5d ago against content hash a223eaf76229, method: parsed. Prices are Anthropic first-party input rates as of 2026-09-05, from the pricing page.

Security

Grade A, and why

shared-language 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 5d 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.

opencode/skills/shared-language/SKILL.md · 59 lines

What it actually says

Shared Language

Build and maintain .opencode/CONTEXT.md — a glossary of project-specific terms, each defined in ONE sentence. It is a glossary and nothing else: no design docs, no ADRs, no examples. If an entry grows beyond one sentence, move it to a design doc and keep a one-sentence pointer here. Every conversation that re-explains the same concept wastes tokens; the glossary replaces those explanations with a single reference.

When to Create or Update

  1. You find yourself explaining the same concept (>2 sentences) across multiple messages or sessions.
  2. You encounter a term the user uses that has project-specific meaning.
  3. After a design discussion — capture the decisions as shared language entries.

Format

.opencode/CONTEXT.md uses this exact format:

# Shared Language

- `<term>`: <one-sentence definition>
- `<term>`: <one-sentence definition>

Each entry is one line. No paragraphs, no examples, no "further reading".

How Agents Use It

Before reasoning about the codebase, scan .opencode/CONTEXT.md:

  1. If a term appears, use its one-sentence definition directly — do not re-explain or expand.
  2. If a concept is missing but you needed 3+ sentences to convey it, add an entry after the session.

Each definition replaces 5-20 sentences of repeated explanation, so the agent spends fewer tokens re-deriving concepts.

ADR triage

Offer an ADR only when a decision is ALL of: hard-to-reverse, surprising (a reader would not guess it), and a real trade-off (both options cost something). Otherwise a .opencode/CONTEXT.md entry suffices — do not spawn an ADR.

Maintenance

  • Keep entries to one sentence; an overgrown entry belongs in a design doc.
  • Delete entries that no longer apply (removed features, renamed concepts).
  • Sort alphabetically.
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. 5d ago First seen · 59 lines · 41 tokens per session scan A a223eaf76229

Subscribe to this mod's changes

shared-language is a skill published in the GitHub repository znlgis/my-opencode-deepseek-config (57 stars, last pushed 2d ago), licensed MIT. It adds 41 tokens to every session and 477 once invoked, about $0.0002 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.