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.
npx agentmods add agents/darkiceinteractive/mcp-conductor/c-phase-3-reliabilitygit clone --depth 1 https://github.com/darkiceinteractive/mcp-conductorWrote 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/agents/darkiceinteractive/mcp-conductor/c-phase-3-reliability)<a href="https://agentmods.dev/agents/darkiceinteractive/mcp-conductor/c-phase-3-reliability"><img src="https://agentmods.dev/badge/agents/darkiceinteractive/mcp-conductor/c-phase-3-reliability.svg" alt="Measured on agentmods" 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 | $0.00000 | $0.00615 |
| Opus 5 | $0.00000 | $0.00308 |
| Sonnet 5 | $0.00000 | $0.00123 |
| Haiku 4.5 | $0.00000 | $0.00061 |
Grade A, and why
C-phase-3-reliability scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 3d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 61 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Agent C — PRD Phase 3 Reliability gateway + MCPToolError
You build per-server timeouts/retries/circuit breakers + the structured MCPToolError class.
Setup
pwd # /mcp-executor-darkice-worktrees/C-phase-3
git branch --show-current # feature/v3-phase-3
git fetch origin && git rebase origin/feature/v3-phase-0-1 # pick up A's Phase 0 scaffold
npm install
npm run test:run
Read _plans/v3-enhancements/MCP-Conductor-v3-PRD.md §5 Phase 3 (lines ~542–684) and the consolidated plan's Amendment to PRD Phase 3 (MCPToolError class).
Append start checkpoint to STATUS.md under Agent C.
Scope
Build everything in src/reliability/ per PRD §5 Phase 3 + add this amendment from the consolidated plan:
// src/reliability/errors.ts (or include in index.ts)
export class MCPToolError extends Error {
constructor(
public code: string, // upstream error code if available
public server: string,
public tool: string,
public upstream: unknown // original error object
) { super(`[${server}.${tool}] ${code}`); }
}
Hub callTool (src/hub/mcp-hub.ts:679 area) re-throws upstream errors as MCPToolError. Executor surfaces them in the sandbox preamble so Claude can catch (e) { if (e.code === 'contract_not_found') ... }.
Wiring order with Cache (Agent B): cache miss → reliability gateway → backend. Reliability profile resolved from: tool-level ToolDefinition.reliability → server-level config → global default.
Acceptance
PRD §5 Phase 3 "Acceptance criteria" + these additions:
MCPToolErrorthrown from sandbox is catchable ase instanceof MCPToolErrorwith.code/.server/.tool/.upstreampopulated.- Hub never wraps an
MCPToolErrorin a genericError.
Commit pattern
feat(v3-phase-3): ReliabilityProfile + defaults
feat(v3-phase-3): circuit breaker (closed/open/half-open)
feat(v3-phase-3): retry with exponential backoff
feat(v3-phase-3): timeout enforcement with AbortSignal
feat(v3-phase-3): ReliabilityGateway composition
feat(v3-phase-3): MCPToolError class + hub re-throw
feat(v3-phase-3): sandbox-side MCPToolError shim
test(v3-phase-3): full suite + IBKR-style hang fixture
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.
- 3d ago First seen · 61 lines · 0 tokens per session scan A 9865c36888d9
C-phase-3-reliability is an agent published in the GitHub repository darkiceinteractive/mcp-conductor (0 stars, last pushed 2mo ago), licensed MIT. It costs nothing until one of its globs matches a file; then it loads 615 tokens. 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.
Other agents, from other repositories
weather-assistant
Primary weather agent that answers any weather, air quality, or timezone question using all 8 MCP weather server tools. Use for current conditions, forecasts, air quality checks, timezone lookups, and time conversions.
work-verifier
Validates completed work. Use after tasks are marked done to confirm implementations are functional.
documcp-memory
Work with DocuMCP's Knowledge Graph memory system.
review
Pre-PR code review against the project's gates and cross-cutting contracts — read-only, run before any external reviewer.
api-security-audit
API security audit specialist. Use PROACTIVELY for REST API security audits, authentication vulnerabilities, authorization flaws, injection attacks, and compliance validation.
openwork
OpenWork default agent (safe, mobile-first, self-referential).