dsh-mcp-panel: Instructions file for Codex

AGENTS.md

dsh-mcp-panel AGENTS.md is an instructions file for Codex, OpenCode from PerryLink/dsh-mcp-panel. It costs 2,144 tokens per session, scanned A, original, Apache-2.0.

Repository instructions for a standalone DeepSeek-based harness plugin that displays read-only status from an MCP tool system. MCP is a standard way for software to expose tools to AI assistants.

In plain words
What is it for?
Use them when working on the plugin entry point, status service, shared data schema, upstream status support, or its build and configuration.
Why use it?
They document the plugin contract, shared status data format, project layout, and build rules needed to change it without breaking the host or client.

Instructions file for CodexOpenCode

Written for Codex and OpenCode: the file is AGENTS.md. Also seen: mentions AGENTS.md.

This is PerryLink/dsh-mcp-panel's own configuration. It tells Codex and OpenCode how to work on dsh-mcp-panel itself, so it is not a mod to install elsewhere. Copy it as a starting point and replace the rules that are about this project. Everything dsh-mcp-panel configures →

Reuse

Borrowing it

Nothing to install: this file belongs to PerryLink/dsh-mcp-panel. Take a copy, put it at the same path in your own repository, and replace the rules that are about this project with yours.

Copy the file
curl -O https://raw.githubusercontent.com/PerryLink/dsh-mcp-panel/main/AGENTS.md
Clone the repo
git clone --depth 1 https://github.com/PerryLink/dsh-mcp-panel

Made for: Codex, OpenCode.

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 dsh-mcp-panel AGENTS.md

README.md
[![agentmods](https://agentmods.dev/badge/instructions/perrylink/dsh-mcp-panel/agents-md/github.svg)](https://agentmods.dev/instructions/perrylink/dsh-mcp-panel/agents-md)
Your own site
<a href="https://agentmods.dev/instructions/perrylink/dsh-mcp-panel/agents-md"><img src="https://agentmods.dev/badge/instructions/perrylink/dsh-mcp-panel/agents-md/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 dsh-mcp-panel AGENTS.md

Your own site · 80×15
<a href="https://agentmods.dev/instructions/perrylink/dsh-mcp-panel/agents-md"><img src="https://agentmods.dev/badge/instructions/perrylink/dsh-mcp-panel/agents-md.svg" alt="Reviewed on agentmods" width="80" height="20"></a>
Per session 2,144 This file is loaded in full into every session.
When invoked 2,144 The same file — it is already loaded in full.
Security scan A 0 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 $0.02144 $0.02144
Opus 5.5 $0.00858 $0.00858
Sonnet 5.5 $0.00429 $0.00429
Haiku 4.5 $0.00214 $0.00214

Measured 1mo ago against content hash 3f4f03ab43f9, method: parsed. Prices are Anthropic first-party input rates as of 2026-10-07, from the pricing page.

Security

Grade A, and why

dsh-mcp-panel AGENTS.md 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.

AGENTS.md · 54 lines

How it starts

The opening of the file, as written. The whole thing — 54 lines — stays where its author put it; the contents beside it link to each section on GitHub.

AGENTS.md

Standalone DeepSeek Harness plugin repository (dsh-mcp-panel). Development follows the dsh-plugin-guide skill and the official plugin contract; this file records repo-local decisions.

Layout

  • src/index.ts — function-plugin contract (name/inject/Config/apply; NO default export — the Loader unwraps exports.default ?? exports).
  • src/service.ts — McpPanelService (TypertRemoteService, namespace mcpPanel): read-only snapshot assembly from loader rows + tool registry + upstream status observations. Serves mcpPanel/status, starts panel-only probes through mcpPanel/probe (probe(serverName); needs ctx.jobs, streamable-http and stdio rows), appends CRUD fragments through mcpPanel/writePatch (approval-gated, backup-first, and RE-VERIFIED against the loader's re-applied state — a skipped patch never reads as success; writeVerifyEnabled/writeVerifyTimeoutMs tune the gate), and trials tools (including the three shared resource tools) through mcpPanel/callTool. Resources capability = ctx.mcpResources present AND the three shared tools registered in this scope.
  • src/wire.ts — the snapshot vocabulary, its zod v4 wire schema, and the single mcpPanel/status invocation descriptor shared verbatim by the host ./typert manifest (src/typert.host.ts) and the client Remote contribution (src/client/remote.ts) — one canonical source so the two codecs can never drift.
  • src/upstream.ts — the proposed upstream mcp/status seam (event + query service face), declared here via cordis declaration merging and consumed with feature detection; when upstream ships it, its identical declarations merge cleanly and a conflicting signature fails this compile (intended tripwire). Proposal text: docs/upstream-proposal.md in the deepseek-harness repo. Resources are bridged by the SHIPPED @deepseek-ai/dsh-mcp-resources service; this file declares its minimal structural face (mcpResources) for presence detection only — resource trials go through the OFFICIAL shared tools, never provider internals.
  • src/sanitize.ts — display redaction (URL query credentials, userinfo passwords, header values, bearer tokens, JWTs). Pure; extreme-case tests in tests/sanitize.spec.ts.
  • src/grouping.ts / src/aggregate.ts — pure enumeration/grouping and status aggregation with missing-field tolerance.
  • src/command.ts — the /mcp command (standard CommandResult; logged via command/run + command/done).
  • src/patch.ts — pure module: config validation against the official client's face, keep-semantics merging, and the loader-dialect fragment renderer. CRUD emits insert (add) and id-targeted overrides (- id: + name: + disabled:/config:) — the harness patch dialect has NO set and no remove, so disabling a row IS the canonical removal (issue #27).
  • src/trial.ts — the official-pipeline trial caller: mcp__* tools plus the three shared resource tools list_mcp_resources / list_mcp_resource_templates / read_mcp_resource, with the requested server injected into their arguments (conflicts fail closed).
  • src/probe.ts — optional mcp_probe background-job tool (unowned job: panel-only results). Both transports probe: streamable-http via one initialize POST; stdio via spawning the row's command/args under scrubbedParentEnv (the same base the mcp-client bridge uses, imported from @deepseek-ai/dsh-subprocess) + explicit env/cwd, one initialize handshake over stdin/stdout, sanitized serverInfo or failure detail.
  • src/client/ — browser half: $mount the Remote contribution, register the settings.plugins.tab entry id mcp, pure presenter in present.ts, inline scoped stylesheet in styles.ts (standalone bundles cannot use the in-repo CSS-module pipeline).
  • tests/ — vitest; REAL Context + Session/ToolRuntime/CommandRuntime from the 0.1.7-rc.2 peers, fake Loader face, fake Agent, optional fake jobs.

Read the full file on GitHub · 54 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. 1mo ago Changed · +3 tokens per session 3f4f03ab43f9
  2. 1mo ago Changed · +51 tokens per session 92cfacacc191
  3. 1mo ago Changed · +2 lines · +338 tokens per session 6242f0f98443
  4. 1mo ago Changed · +1 tokens per session f5caf278d987
  5. 2mo ago Changed · +7 tokens per session 24f19e7c216d
  6. 2mo ago Changed · -4 tokens per session 3afffe0eb09f
  7. 2mo ago First seen · 52 lines · 1,748 tokens per session scan A c0d5af43fda4

Subscribe to this mod's changes

dsh-mcp-panel AGENTS.md is an instructions file published in the GitHub repository PerryLink/dsh-mcp-panel (93 stars, last pushed 2d ago), licensed Apache-2.0. It adds 2,144 tokens to every session, about $0.0086 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 2026-08-30.

Related

dsh-mcp-panel AGENTS.md compared with instructions from other repositories

dsh-mcp-panel AGENTS.md next to 6 instructions from other repositories
ModTypeStarsLast pushLicenceScanTokensWorks with
dsh-mcp-panel AGENTS.mdPerryLink/dsh-mcp-panelInstructions file932d agoApache-2.0A2,144Codex, OpenCode
vscode buildNext.instructions.mdmicrosoft/vscodeInstructions file195ktodayMITA9,719GitHub Copilot
vscode oss-third-party-notices.instructions.mdmicrosoft/vscodeInstructions file195ktodayMITA5,516GitHub Copilot
next.js AGENTS.mdvercel/next.jsInstructions file143kyesterdayMITA7,461Codex, OpenCode
langchain AGENTS.mdlangchain-ai/langchainInstructions file149ktodayMITA4,680Codex, OpenCode
vscode copilot-instructions.mdmicrosoft/vscodeInstructions file195ktodayMITA3,093GitHub Copilot
vscode model-prompts.instructions.mdmicrosoft/vscodeInstructions file195ktodayMITA2,022GitHub Copilot