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/k-integrationgit clone --depth 1 https://github.com/darkiceinteractive/mcp-conductorWhat 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 | $0.00980 |
| Opus 5 | $0.00000 | $0.00490 |
| Sonnet 5 | $0.00000 | $0.00196 |
| Haiku 4.5 | $0.00000 | $0.00098 |
Grade A, and why
K-integration 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.
How it starts
The opening of the file, as written. The whole thing — 88 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Agent K — Integration day
You're the final agent. All 10 prior PRs are merged to feature/v3 (or about to be). You run benchmarks, soak test, finish docs, and tag v3.0.0-beta.1.
Setup
You run at the main repo path (no worktree). The other worktrees may still exist; that's fine — they'll be cleaned up after.
cd /Users/mattcrombie/Dev/Projects/Claude/mcp-executor-darkice
git checkout feature/v3
git pull
git log --oneline -30 # confirm all 10 PRs are in
npm install && npm run build && npm run test:run && npm run lint
Read STATUS.md to see what each agent reported. Confirm all 10 sections show READY-FOR-MERGE or completed.
Append start checkpoint to STATUS.md under Agent K.
Scope
Block 1 — Merge any open PRs in dependency order
A → (B, C, D, G, H) parallel → (E, F) → (I after F, J after H)
Use gh pr merge --squash --auto if CI is green. If conflicts: rebase the lagging branch on feature/v3 and re-push.
Block 2 — Head-to-head benchmark vs Anthropic 150K → 2K
Implement the Google-Drive-to-Salesforce style chain as a benchmark fixture:
- New file:
test/benchmark/anthropic-pattern.test.ts - New fixture:
test/fixtures/google-drive-to-salesforce.ts(synthetic 150K-token document chain → record extraction → Salesforce update) - Assert ≥ 98% reduction (matches Anthropic's claim)
- Output:
docs/benchmarks/anthropic-comparison-YYYY-MM-DD.{json,md}
This backs the positioning claim "the production implementation of Anthropic's published design."
Block 3 — One-hour soak test against fault-injected backends
Use Phase 7's replay infrastructure. Synthetic faults: 10% timeout, 5% 500-error, 1% truncated response. Assert 0 hangs, all errors surfaced as MCPToolError.
Block 4 — Complete docs
Per PRD §6.4:
docs/v3/architecture.md— diagrams + component overviewdocs/v3/migration.md— v2 → v3 migrationdocs/v3/configuration.md— full config reference (registry/cache/reliability/runtime/skills/findTool/daemon/observability sections)docs/v3/sandbox-api.md— updatedmcpAPI (helpers, skills, shared, tokenize, fs)docs/v3/recipes.md— example execute_code workflows showcasing v3 features- Update top-level
README.mdwith v3 highlights
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.
- yesterday First seen · 88 lines · 0 tokens per session scan A 394259fa6606
K-integration 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 980 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
git-agent
Background git operations agent - commits, PRs, branch management, release workflows. Runs on Sonnet to free main session.
release
Automate the release workflow — bump version, build, publish to npm, commit, tag, and push. Use when asked to cut a release or publish a new version.
ecosystem-changelog-ecosystem
Status: human guide. Owner: Emerson Braun. Reviewed: 2026-08-14.
deployment-manager
Deployment and release management specialist. Handles npm publishing, Docker containerization, version management, and production deployments. Use PROACTIVELY before releases.
changelog-generator
Generates changelog from commits.
fakoli-sentinel
Fakoli Crew QA validator for tests, validation scorecards, release readiness, and evidence-backed pass/fail calls.