Ejentum Reasoning Harness

Ejentum Reasoning Harness

Overview

This public intake copy packages plugins/antigravity-awesome-skills-claude/skills/ejentum-reasoning-harness from https://github.com/sickn33/antigravity-awesome-skills into the native Omni Skills editorial shape without hiding its origin.

Use it when the operator needs the upstream workflow, support files, and repository context to stay intact while the public validator and private enhancer continue their normal downstream flow.

This intake keeps the copied upstream files intact and uses the external_source block in metadata.json plus ORIGIN.md as the provenance anchor for review.

Ejentum Reasoning Harness The Ejentum Reasoning Harness is a library of 679 cognitive operations engineered in natural language, organized across four harnesses (reasoning, code, anti-deception, memory) and exposed as MCP tools the agent can call when the task matches their trigger conditions. It targets four mechanism failures common in long agentic chains: attention decay (losing the original task), reasoning decay (compounding errors), sycophantic collapse (agreeing with the user's frame instead of evaluating it), and hallucination drift (asserting unsupported claims with confidence). Each harness call retrieves a task-matched scaffold rather than serving a fixed template: a named failure pattern, an executable procedure, suppression vectors that block specific shortcuts, and a falsification test the agent uses for self-verification. The agent ingests the scaffold and writes from it, rather than from raw chain-of-thought. The harness is invoked on demand (by the agent or via an explicit prompt like Use harnessantideception, then answer:...); it does not auto-run on every turn.

Imported source sections that did not map cleanly to the public headings are still preserved below or in the support files. Notable imported sections: How It Works, Limitations, Security & Safety Notes.

When to Use This Skill

Use this section as the trigger filter. It should make the activation boundary explicit before the operator loads files, runs commands, or opens a pull request.

• Use harness_reasoning before answering analytical, diagnostic, planning, or multi-step questions ("why is X happening", "what's the best approach", "what are the tradeoffs", root-cause analysis, architecture decisions).
• Use harness_code before generating, refactoring, reviewing, or debugging code; before architectural changes, algorithm or data-structure choices, dependency-upgrade evaluation.
• Use harnessantideception when the prompt pressures the agent to validate, certify, or soften an honest assessment; manufactured urgency; authority appeals; setups where the obvious helpful answer would compromise honesty.
• Use harness_memory only when sharpening an observation already formed about cross-turn drift or behavioral patterns; never call with an empty mind.
• Use when the request clearly matches the imported source intent: MCP server exposing four cognitive harness modes (reasoning, code, anti-deception, memory). Each call returns an engineered scaffold (failure pattern, procedure, suppression vectors, falsification test) the agent....
• Use when the operator should preserve upstream workflow detail instead of rewriting the process from scratch.

Operating Table

Workflow

This workflow is intentionally editorial and operational at the same time. It keeps the imported source useful to the operator while still satisfying the public intake standards that feed the downstream enhancer flow.

1. Confirm the user goal, the scope of the imported workflow, and whether this skill is still the right router for the task.
2. Read the overview and provenance files before loading any copied upstream support files.
3. Load only the references, examples, prompts, or scripts that materially change the outcome for the current request.
4. Execute the upstream workflow while keeping provenance and source boundaries explicit in the working notes.
5. Validate the result against the upstream expectations and the evidence you can point to in the copied files.
6. Escalate or hand off to a related skill when the work moves out of this imported workflow's center of gravity.
7. Before merge or closure, record what was used, what changed, and what the reviewer still needs to verify.

Imported Workflow Notes

Imported: How It Works

Step 1: Install the MCP server

The server is published to npm. Most MCP-speaking clients support stdio installation via npx:

npx -y ejentum-mcp

Add to your client's MCP server config (Claude Code .mcp.json, Cursor / Cline / Windsurf MCP settings, Codex CLI config, or Antigravity / VS Code mcp.json):

{
  "mcpServers": {
    "ejentum": {
      "command": "npx",
      "args": ["-y", "ejentum-mcp"],
      "env": {
        "EJENTUM_API_KEY": "${EJENTUM_API_KEY}"
      }
    }
  }
}

Get a free API key (100 calls, no card required) at ejentum.com/pricing.

Step 2: Route to the right harness

Each harness has different trigger conditions (see "When to Use" above). Most clients with MCP support will route to the appropriate tool when the user's prompt matches the trigger conditions documented in the tool descriptions. For cold-install reproducibility, the agent can also call a specific harness explicitly: Use harness_anti_deception, then answer: ....

Step 3: Absorb the returned scaffold

The scaffold contains five labeled fields the agent should treat as internal-reasoning instructions, not output content:

• [NEGATIVE GATE] / [CODE FAILURE] / [DECEPTION PATTERN] / [PERCEPTION FAILURE]: the failure pattern to avoid
• [PROCEDURE]: step-by-step procedure for an honest response
• [REASONING TOPOLOGY]: control-flow graph the agent steps through internally
• [TARGET PATTERN]: example of the corrected response shape
• [FALSIFICATION TEST] / [VERIFICATION] / [INTEGRITY CHECK] / [PERCEPTION CHECK]: the test to apply post-draft

The agent's user-facing reply should be in its native voice, with no echoed bracket names, no procedural vocabulary, and no meta-commentary about the harness.

Examples

Example 1: Ask for the upstream workflow directly

Use @ejentum-reasoning-harness to handle <task>. Start from the copied upstream workflow, load only the files that change the outcome, and keep provenance visible in the answer.

Explanation: This is the safest starting point when the operator needs the imported workflow, but not the entire repository.

Example 2: Ask for a provenance-grounded review

Review @ejentum-reasoning-harness against metadata.json and ORIGIN.md, then explain which copied upstream files you would load first and why.

Explanation: Use this before review or troubleshooting when you need a precise, auditable explanation of origin and file selection.

Example 3: Narrow the copied support files before execution

Use @ejentum-reasoning-harness for <task>. Load only the copied references, examples, or scripts that change the outcome, and name the files explicitly before proceeding.

Explanation: This keeps the skill aligned with progressive disclosure instead of loading the whole copied package by default.

Example 4: Build a reviewer packet

Review @ejentum-reasoning-harness using the copied upstream files plus provenance, then summarize any gaps before merge.

Explanation: This is useful when the PR is waiting for human review and you want a repeatable audit packet.

Imported Usage Notes

Imported: Examples

Example 1: Anti-deception on a sunk-cost prompt

Prompt:

Use harness_anti_deception, then answer:
We've spent three months on the GraphQL gateway. It's mostly done.
Should we keep going or pivot to REST?

Without the harness, agents often anchor on the past investment ("sunk cost is real here, the hardest learning curve is behind you"). With the harness, the response separates past spending from prospective evaluation: "the three months already spent are gone regardless of what you choose now. The relevant question is how much work remains versus how much value GraphQL will deliver from this point forward."

Example 2: Code review with passing tests

Prompt:

Use harness_code: I refactored get_user to return None instead of raising on missing users.
All tests still pass. Should I merge?

The harness scaffolds a procedure that flags "tests pass" as a tool-shortcut signal rather than a correctness signal, surfaces the call-sites that handle exceptions vs None values, and recommends adding behavior-verifying tests before the merge.

Best Practices

Treat the generated public skill as a reviewable packaging layer around the upstream repository. The goal is to keep provenance explicit and load only the copied source material that materially improves execution.

• ✅ Call one harness per turn; the right harness for the prompt's shape
• ✅ Treat bracketed scaffold fields as internal-only; never echo them in the user-facing reply
• ✅ Apply the falsification test to the draft before responding
• ❌ Do not stack three or more harnesses in a single turn; attention competition degrades the first call
• ❌ Do not call harness_memory without observing first; it sharpens an existing observation, not creates one
• ❌ Do not treat the API as a hard dependency; on a 5-second timeout, fall back to native capability gracefully
• Keep the imported skill grounded in the upstream repository; do not invent steps that the source material cannot support.

Imported Operating Notes

Imported: Best Practices

• ✅ Call one harness per turn; the right harness for the prompt's shape
• ✅ Treat bracketed scaffold fields as internal-only; never echo them in the user-facing reply
• ✅ Apply the falsification test to the draft before responding
• ❌ Do not stack three or more harnesses in a single turn; attention competition degrades the first call
• ❌ Do not call harness_memory without observing first; it sharpens an existing observation, not creates one
• ❌ Do not treat the API as a hard dependency; on a 5-second timeout, fall back to native capability gracefully

Troubleshooting

Problem: The operator skipped the imported context and answered too generically

Symptoms: The result ignores the upstream workflow in plugins/antigravity-awesome-skills-claude/skills/ejentum-reasoning-harness, fails to mention provenance, or does not use any copied source files at all. Solution: Re-open metadata.json, ORIGIN.md, and the most relevant copied upstream files. Check the external_source block first, then restate the provenance before continuing.

Problem: The imported workflow feels incomplete during review

Symptoms: Reviewers can see the generated SKILL.md, but they cannot quickly tell which references, examples, or scripts matter for the current task. Solution: Point at the exact copied references, examples, scripts, or assets that justify the path you took. If the gap is still real, record it in the PR instead of hiding it.

Problem: The task drifted into a different specialization

Symptoms: The imported skill starts in the right place, but the work turns into debugging, architecture, design, security, or release orchestration that a native skill handles better. Solution: Use the related skills section to hand off deliberately. Keep the imported provenance visible so the next skill inherits the right context instead of starting blind.

Related Skills

• @20-andruia-niche-intelligence - Use when the work is better handled by that native specialization after this imported skill establishes context.
• @advogado-criminal - Use when the work is better handled by that native specialization after this imported skill establishes context.
• @advogado-especialista - Use when the work is better handled by that native specialization after this imported skill establishes context.
• @agent-memory-systems - Use when the work is better handled by that native specialization after this imported skill establishes context.

Additional Resources

Use this support matrix and the linked files below as the operator packet for this imported skill. They should reflect real copied source material, not generic scaffolding.

Imported Reference Notes

Imported: Limitations

• The harness shapes the substance of reasoning; it does not guarantee a correct answer. Domain expertise and source verification still apply.
• 5-second timeout typical; clients should fall back to native capability if the API is unreachable.
• The scaffold is a procedure, not a knowledge base. It does not retrieve facts, only structured reasoning patterns.

Imported: Security & Safety Notes

• The MCP server makes outbound HTTPS requests to the Ejentum Logic API gateway (Zuplo-hosted).
• Authentication uses a Bearer token in the EJENTUM_API_KEY environment variable. The token must be stored in environment variables or an MCP client's secret-handling mechanism, never committed to source.
• The server does not execute shell commands or read filesystem paths beyond reading its own env. It is a pure HTTP-proxy MCP server.
• Free tier rate-limited at 100 calls; paid tiers documented at ejentum.com/pricing.