add-target

A maintenance workflow for stdagent's list of supported AI coding tools and their project configuration formats. A target is one supported tool or platform.

In plain words
What is it for?
Use it to scan for new platforms, verify existing targets, update adapters and documentation, or prepare unsupported targets for removal.
Why use it?
It reduces the risk of supporting outdated tools, missing newly available project settings, or changing adapters without checking official documentation.

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/stringke/std-agent/add-target
Any agent
npx skills add StringKe/std-agent --skill add-target
Clone the repo
git clone --depth 1 https://github.com/StringKe/std-agent

Made for: Claude Code, Codex.

Per session 71 Skills are progressive disclosure: only the name and description are preloaded; the body loads when the skill is used.
When invoked 1,600 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 $0.00071 $0.01600
Opus 5 $0.00036 $0.00800
Sonnet 5 $0.00014 $0.00320
Haiku 4.5 $0.00007 $0.00160

Measured yesterday against content hash 5f82c06eb1c5, method: parsed. Prices are Anthropic first-party input rates as of 2026-08-30, from the pricing page.

Security

Grade A, and why

add-target 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 yesterday.

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.

.grok/skills/add-target/SKILL.md · 91 lines

How it starts

The opening of the file, as written. The whole thing — 91 lines — stays where its author put it; the contents beside it link to each section on GitHub.

Maintain targets

目标:ValidTargets 与仍在维护、且有项目级配置面的 AI CLI 对齐。新平台按官方能力接入 5 种 type;已有平台只改有证据的差异;官方停更或配置面消失则移除。不破坏共享输出。

开始前读取 .stdai/standards/references/transformer-design.mddocs/format-spec.mddocs/conversion-rules.md,以及相近 target 的 docs/targets/<name>.md 与 transformer。

  • 用户点名工具:只处理该 target(含接入、更新或移除)。
  • 未点名:必须同时做三件事,先出差异表再改代码。无证据变化不要重写。
    1. 发现未注册的新平台
    2. 复核每个 ValidTargets 的官方规范
    3. 判定停更项并准备移除

未要求时不 commit、不发版。移除是 breaking change:先列出官方证据和受影响路径,确认后再删代码。

发现新平台

对照 internal/config.ValidTargets 搜索尚未接入的 AI coding CLI / IDE agent。用 exa 查产品与新闻,用 context7 和官方站点查项目级配置,用 gh_grep 核仓库实现。社区文章只作明确标注的补充。

接入门槛:有文档化的项目级 rules/skills/commands/agents 配置,且 stdagent 能写出该工具会发现的文件。只有用户级 ~/ 配置或纯聊天 UI 的列入 skip,不要接入。

每个候选记 add / skip / watch,并带来源 URL。

调研现有 target

官方文档和官方仓库是协议依据。每条结论带完整来源 URL。

用当前官方页面对照 adapter 与 docs/targets/<name>.md。逐项标 native / degrade / skip / UNKNOWN:

维度 要确认的事实
存活 产品是否仍发布;仓库是否 archived;是否改名/被后继产品取代
发现路径 项目根文件、目录、加载顺序、嵌套、本地覆盖文件
rules 原生路径、path matcher、always-on、是否 inline 到根文件
skills 是否 Agent Skills;目录;frontmatter 白名单
commands 原生 prompts/commands/workflows,或降为 skill 子目录
references 原生目录或 fallback
subagents 原生 agents 目录与字段,或 fallback
MCP 项目级文件与顶层 key;用户配置合并则不要覆盖
共享路径 是否消费 AGENTS.md / .agents/;多 target 同路径必须字节一致
限额 单文件、根文件、总量的 soft/hard;超限是截断、拒载还是淘汰
不写 hooks、用户 settings、全局 ~/ 配置、crush.json / kilo.jsonc 这类 JSONMerge 用户文件

硬限制只记录有官方证据的值,并同步 internal/budget.Limits。没有文档化上限就写 UNKNOWN,不要发明数字。

停更与移除

满足任一官方证据则移除(或先迁到后继 target):

  • 官方宣布停更、sunset 或产品下线
  • 官方仓库 archived,且没有沿用同一项目配置的后继产品
  • 项目级配置面消失,stdagent 再也写不出该工具会发现的文件
  • 产品改名或被后继取代:优先把现有 target 迁到后继约定,不要留僵尸名

不要因为发布间隔长、用户少、文档暂时 404 就移除。证据不足标 UNKNOWN 并保留。

改名迁移不算静默删除:保留用户仍在用的路径,直到后继 adapter 能写出等价文件。

ValidTargets 删除会使旧 config.toml 校验失败,commit 使用 !BREAKING CHANGE。共享路径(如 AGENTS.md)只要还有其他 producer 就保留。本仓库先从 .stdai/config.toml 去掉该 target,再 go run ./cmd/stdagent sync --no-pull --strict 让 prune 清掉其独占生成物,然后删代码。

Read the full file on GitHub · 91 lines

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. yesterday First seen · 91 lines · 71 tokens per session scan A 5f82c06eb1c5

Subscribe to this mod's changes

add-target is a skill published in the GitHub repository StringKe/std-agent (8 stars, last pushed 11d ago), licensed MIT. It adds 71 tokens to every session and 1,600 once invoked, about $0.0004 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 2026-08-31.