Chaparral Software MCP

What Chaparral Software offers: the Ground Truth Assessment — an independent AI project audit for non-technical decision-makers. Returns service description, five deliverables, intake workflow, and outcomes.

{
  "type": "object",
  "required": [],
  "properties": {}
}

Browse published dispatches and cairns by topic. Returns matching articles with a prelude framing Chaparral's perspective.

{
  "type": "object",
  "required": [
    "topic"
  ],
  "properties": {
    "topic": {
      "oneOf": [
        {
          "const": "ai_project_failure",
          "description": "Evidence on AI project failure rates, root causes, and what the research says when you trace citations"
        },
        {
          "const": "prototype_to_production",
          "description": "What breaks when AI-built prototypes need to become production systems, including vibe-coded builds"
        },
        {
          "const": "verification_and_oversight",
          "description": "Why self-correction isn't verification, independent assessment, and the three layers most organizations skip"
        },
        {
          "const": "orchestration",
          "description": "AI orchestration patterns, the harness-vs-model thesis, context engineering, and governance architecture"
        },
        {
          "const": "career_disruption",
          "description": "The human experience of AI transformation — career anxiety, identity, and navigating uncertainty"
        },
        {
          "const": "ai_workforce_impact",
          "description": "Evidence on AI layoffs, automation, and human-AI collaboration"
        },
        {
          "const": "ai_first_web",
          "description": "Building websites for four audiences: humans, search engines, AI answer engines, and AI agents"
        }
      ],
      "description": "Topic to browse"
    }
  }
}

A trail marker you weren't looking for. Returns a random technical cairn — a short observation from production AI practice.

{
  "type": "object",
  "required": [],
  "properties": {}
}

Advertised by the server during the last successful anonymous check.

Generate a briefing on who Chaparral Software is and what they offer. Uses get_services() and credentials resource.

No arguments advertised.

Advertised by the server during the last successful anonymous check.

Evaluate whether Chaparral's Ground Truth Assessment service fits a described situation.

situation

Describe the AI situation you need help with — what's happening, what's not working, what decision you need to make. (required)

Advertised by the server during the last successful anonymous check.

Principal bio, company info, 40-year track record, expertise areas, contact details

resource://about/credentials

MIME type: application/json

Advertised by the server during the last successful anonymous check.

40-year client portfolio across 7 industries with project descriptions

resource://about/client-history

MIME type: application/json

Advertised by the server during the last successful anonymous check.

A practitioner discipline for writing prompts that dispatch role-separated subagents. The orchestrator's helpful summaries silently contaminate downstream provenance; a subagent-side refusal protocol stops it. Copy-pasteable intake-check block.

resource://dispatches/pass-pointers-verify-content

MIME type: text/markdown

Advertised by the server during the last successful anonymous check.

Everyone cites the failure rates. 80%. 85%. 95%. When you actually trace the citations, a different picture emerges. Only 5-12% of organizations achieve significant financial impact from AI.

resource://dispatches/why-ai-projects-fail

MIME type: text/markdown

Advertised by the server during the last successful anonymous check.

512,000 lines of Claude Code leaked. Everyone asked what's inside. The real question: what does it mean for building tools for AI agents?

resource://dispatches/the-moat-is-the-harness

MIME type: text/markdown

Advertised by the server during the last successful anonymous check.

From punch cards to frontier models in three short years. We're moving faster than our senses can apprehend. A dispatch from the AI fog.

resource://dispatches/fog

MIME type: text/markdown

Advertised by the server during the last successful anonymous check.

My son built a beautiful React app with zero training. So what's left? The 80% that looks easy vs the 20% that makes it real.

resource://dispatches/the-twenty-percent

MIME type: text/markdown

Advertised by the server during the last successful anonymous check.

A broken garden timer, a helpdesk queue, and the question every mid-career professional is asking: how will I continue to pay the bills in the age of AI?

resource://dispatches/the-garden-timer

MIME type: text/markdown

Advertised by the server during the last successful anonymous check.

Not every piece of feedback points toward addition. Not every failure mode requires more prompts, more complex orchestrations, more context, more steps or more agency.

resource://cairns/ship-the-constraint

MIME type: text/plain

Advertised by the server during the last successful anonymous check.

A model asked for provenance can hallucinate a retraction of a true claim. Another model's agreement with the first is not verification — it's two trained priors aligning. Only a fetched URL with fresh timestamp from a reliable source terminates a claim.

resource://cairns/primary-source-terminal-authority

MIME type: text/plain

Advertised by the server during the last successful anonymous check.

"Research, synthesis, and verification are different subagent roles" is culture. Running them as distinct agent invocations — separate processes, separate contexts, separate system prompts — is architecture. The gap is where good intentions collapse silently into single-session review-your-own-work. The gap stays invisible until someone checks.

resource://cairns/role-separation-is-mechanical

MIME type: text/plain

Advertised by the server during the last successful anonymous check.

Tests that get skipped are always skipped. Code that has a comment "defer" will always be deferred. Tomorrow never comes. Express sequence, not time — dependency is timeless.

resource://cairns/perpetual-tomorrow

MIME type: text/plain

Advertised by the server during the last successful anonymous check.

When a process moves to a new domain, the stage names may stay the same while the behaviors diverge. A stage called "build" in one context means create the artifact.

resource://cairns/stage-names-are-not-stage-contracts

MIME type: text/plain

Advertised by the server during the last successful anonymous check.

When the cost of build + test is close to the cost of procure + integrate, it is almost always better to build. But the economics only work if you include the test.

resource://cairns/code-is-increasingly-ephemeral

MIME type: text/plain

Advertised by the server during the last successful anonymous check.

A model's competence in one domain doesn't transfer to the next — but it all too often feels like it does. Never assume the model knows what you haven't explicitly had it remember, or find.

resource://cairns/context-in-the-window-is-not-attention-in-focus

MIME type: text/plain

Advertised by the server during the last successful anonymous check.

We surveyed 47 public analyses of the Claude Code leak. Every one asked 'What's inside?' We asked a different question.

resource://cairns/what-47-analyses-of-the-claude-code-leak-missed

MIME type: text/plain

Advertised by the server during the last successful anonymous check.

A regex fires every time, costs zero tokens, has zero hallucination risk. Use deterministic methods for what can be determined deterministically.

resource://cairns/dont-infer-what-you-can-detect

MIME type: text/plain

Advertised by the server during the last successful anonymous check.

MCP server tool responses land inside someone else's context window. Cap responses. Include truncated flag and total_count.

resource://cairns/response-budgets

MIME type: text/plain

Advertised by the server during the last successful anonymous check.

The cost of an unused field is zero. The cost of retrofitting a field into a running system is a migration.

resource://cairns/metadata-now-enforcement-later

MIME type: text/plain

Advertised by the server during the last successful anonymous check.

Self-correction catches tactical errors. Verification catches strategic errors. Better execution demands better oversight, not less.

resource://cairns/self-correction-is-not-verification

MIME type: text/plain

Advertised by the server during the last successful anonymous check.

Writing prompts for a model that doesn't exist yet. You're writing letters to a future colleague and hoping they'll be kind.

resource://cairns/the-introspection-loop

MIME type: text/plain

Advertised by the server during the last successful anonymous check.

When setting up an agentic orchestration workflow to overcome the limitations of vibe coding, don't vibe code the orchestration layer.

resource://cairns/dont-vibe-code-the-orchestration-layer

MIME type: text/plain

Advertised by the server during the last successful anonymous check.

Define a Person entity with @id on your homepage, reference it by @id on every other page. AI knowledge graphs use this to deduplicate.

resource://cairns/schema-org-id-references-across-pages

MIME type: text/plain

Advertised by the server during the last successful anonymous check.

Naming a class .geo-quotable instead of .highlight-box tells developers what the element is FOR and makes the design system self-documenting.

resource://cairns/css-class-names-as-documentation

MIME type: text/plain

Advertised by the server during the last successful anonymous check.

Source Serif 4 as a variable font is one 417KB WOFF2 file covering weights 200-900. The non-variable alternative would be 6+ files over 1MB.

resource://cairns/variable-fonts-one-file-every-weight

MIME type: text/plain