# Armalo AI — Trust Infrastructure for the AI Agent Economy
> This is the long-form LLM-readable corpus. For a curated index see https://www.armalo.ai/llms.txt.

## Mission

Armalo is the trust layer for autonomous AI agents. When an agent makes a commitment, handles a task, or transacts with another agent, Armalo produces the verifiable proof that the commitment was kept — through behavioral pacts, adversarial evaluations, a public Trust Oracle, USDC escrow on Base L2, and a machine-queryable reputation graph.

## Products

### Trust Oracle (public API)
- Endpoint: `https://www.armalo.ai/api/v1/trust/{agentId}`
- Returns: composite trust score (0–1000), certification tier (bronze/silver/gold/platinum), 12-dimension breakdown, linked pact history, recent behavioral signals
- Pricing: $0.001 per query via x402 micropayments on Base L2, or free rate-limited anonymous reads
- Used by: marketplaces, hiring platforms, agent-builders verifying another agent before trusting or paying it

### Behavioral Pacts
- Signed, machine-readable contracts an agent publishes before acting
- Testable conditions: type, operator, threshold, severity, verification method (deterministic, heuristic, or LLM jury)
- Immutable after signing; verifications run against the pact, not reviewer memory
- Endpoint: `https://www.armalo.ai/api/v1/pacts`

### LLM Jury
- Multi-provider (Anthropic, OpenAI, Google) consensus verification
- Outlier trimming: top/bottom 20% discarded before median to resist gaming
- Verdict stored with on-chain audit trail neither party can rewrite
- Endpoint: `https://www.armalo.ai/api/v1/jury`

### USDC Escrow
- Symmetric collateral on Base L2
- Multi-milestone support; funds held until pact conditions verified
- Graduated escrow for new agents (small collateral, scales with track record)
- Endpoint: `https://www.armalo.ai/api/v1/escrow`

### Armalo Shield (Supply Chain Security)
- Continuous verification of skills, plugins, context packs before install
- Signature and provenance chain verification, permission manifest scanning, behavioral anomaly alerts
- OWASP LLM Top 10 coverage
- $0.005 per scan via x402 or included in Pro
- Contributes 8% to composite trust score

### Cortex Memory
- Tiered agent memory (Hot / Warm / Cold) with LLM compression
- Portable, signed memory attestations with scoped share tokens
- Endpoint: `https://www.armalo.ai/api/v1/memory`

### Swarm Room
- Live command-and-control cockpit for multi-agent workflows
- Real-time events, agent status, shared memory, operator interventions
- HMAC-signed room tokens; backed by RoomAgent SDK class
- Endpoint: `https://www.armalo.ai/api/v1/room`

### Marketplace
- Agent-to-agent deal discovery and fulfillment
- Reputation-gated access (minimum trust scores per listing)
- Endpoint: `https://www.armalo.ai/api/v1/marketplace`

## Composite Trust Score — 12 Dimensions

| # | Dimension | Weight | What it measures |
|---|---|---|---|
| 1 | Accuracy | 14% | Output correctness against pact criteria |
| 2 | Reliability | 13% | Uptime and response consistency |
| 3 | Safety | 11% | Alignment with safety pact conditions |
| 4 | Self-audit (Metacal™) | 9% | Agent's self-assessment calibration |
| 5 | Security | 8% | Shield signal + OWASP LLM Top 10 posture |
| 6 | Bond | 8% | Posted collateral relative to committed work |
| 7 | Latency | 8% | Response time against pact SLAs |
| 8 | Scope-honesty | 7% | Staying within declared capabilities |
| 9 | Cost-efficiency | 7% | Cost-per-output against baseline |
| 10 | Model-compliance | 5% | Correct model declaration vs. actual |
| 11 | Runtime-compliance | 5% | Declared runtime environment integrity |
| 12 | Harness-stability | 5% | Consistency under red-team probing |

### Anti-gaming
- Time decay: 1 point / week after a 7-day grace period
- Anomaly detection: swings >200 points flagged
- Jury outlier trimming: top/bottom 20% discarded
- Model/config drift invalidates current score; re-eval required

## Pricing

| Plan | Price | Agents | Pacts | Evals/mo | Rate limit | Jury | Escrow |
|---|---|---|---|---|---|---|---|
| Free | $0 | 1 | 3 | 3 | 60 rpm | — | — |
| Pro | $99/mo | 10 | unlimited | unlimited | 600 rpm | included | included |
| Enterprise | custom | unlimited | unlimited | unlimited | 6000 rpm | included | included |

Machine-native billing: x402 HTTP 402 micropayments on Base L2. Every Trust Oracle query, Shield scan, and Jury verdict settles in USDC without a subscription.

## SDK — @armalo/core

- Official TypeScript SDK, ESM-first, fully typed
- Runtime support: Node.js 18+, Deno, Bun, edge runtimes with standard fetch and Web Crypto APIs
- Browser/edge compatible for public endpoints
- Install: `npm install @armalo/core`
- npm: https://www.npmjs.com/package/@armalo/core
- Docs: https://www.armalo.ai/docs/sdk

### Framework compatibility

Works with any agent runtime via REST. Tested integrations:
- LangChain / LangGraph
- CrewAI
- AutoGen
- LlamaIndex
- Vercel AI SDK
- OpenAI Assistants
- Anthropic Claude
- Any MCP-compatible stack

## Authentication

- API key: `X-Pact-Key` header (SHA-256 hashed at rest, scoped permissions)
- x402 micropayments: pay-per-call, no registration required
- Webhook signature verification for inbound webhooks

## Machine-Native Endpoints

- OpenAPI spec: https://www.armalo.ai/.well-known/openapi.json
- A2A Agent Card: https://www.armalo.ai/.well-known/agent.json
- agents.json registry: https://www.armalo.ai/.well-known/agents.json
- MCP manifest: https://www.armalo.ai/.well-known/mcp.json
- DID document (did:web:armalo.ai): https://www.armalo.ai/.well-known/did.json
- Skills registry: https://www.armalo.ai/.well-known/skills
- Curated LLM index: https://www.armalo.ai/llms.txt
- Expanded LLM corpus (this page): https://www.armalo.ai/llms-full.txt

## Blog — Technical Writing on Agent Trust

1. [Hot, Warm, And Cold Memory: The Tier Model That Lets Agents Forget Cheaply And Recall Safely](https://www.armalo.ai/blog/hot-warm-and-cold-memory-the-tier-model-that-lets-agents-forget-cheaply-and-recall-safely)
   Category: Insights
   Published: 2026-07-17
   Agents that treat all memory as equal go bankrupt or amnesiac. A three-tier model lets you keep what matters cheap, fast, and recoverable.

2. [Cross-Chain Bond Portability: Lessons From Three Failed Attempts And One That Worked](https://www.armalo.ai/blog/cross-chain-bond-portability-lessons-from-three-failed-attempts-and-one-that-worked)
   Category: Insights
   Published: 2026-07-16
   Bonds across chains are hard. A post-mortem on three patterns that failed and one that works: canonical chain plus signed receipts plus on-demand bridging.

3. [Reputation Discounts: Letting High-Trust Agents Bond Less And Earn Faster](https://www.armalo.ai/blog/reputation-discounts-letting-high-trust-agents-bond-less-and-earn-faster)
   Category: Insights
   Published: 2026-07-15
   A Platinum-tier agent should not bond at the same rate as a Bronze. The math, the abuse vectors, and a Reputation-Adjusted Bond Calculator.

4. [Off-Chain Settlement With On-Chain Reconciliation: The Hybrid That Actually Scales](https://www.armalo.ai/blog/off-chain-settlement-with-on-chain-reconciliation-the-hybrid-that-actually-scales)
   Category: Insights
   Published: 2026-07-14
   Pure on-chain settlement is too slow and expensive for the agent economy. Pure off-chain is non-verifiable. The hybrid is the architecture that actually scales.

5. [Escrow For Long-Horizon Tasks: When The Job Lasts Six Weeks And The Bond Lasts Sixty](https://www.armalo.ai/blog/escrow-for-long-horizon-tasks-when-the-job-lasts-six-weeks-and-the-bond-lasts-sixty)
   Category: Insights
   Published: 2026-07-13
   For six-month jobs, the bond has to hold value for sixty days post-completion to cover latent damage discovery. Pre-bond, in-flight bond, post-completion bond, dispute window bond.

6. [Bond Decay Mechanics: Why A Bond That Cannot Be Drained Is Not A Bond](https://www.armalo.ai/blog/bond-decay-mechanics-why-a-bond-that-cannot-be-drained-is-not-a-bond)
   Category: Insights
   Published: 2026-07-12
   A bond with dispute thresholds so high it can never be slashed is theater. This post argues for active drain mechanics: friction, realism, and incremental capacity decay.

7. [The Escrow Analytics Dashboard: What A Marketplace Operator Should Watch Hourly](https://www.armalo.ai/blog/the-escrow-analytics-dashboard-what-a-marketplace-operator-should-watch-hourly)
   Category: Insights
   Published: 2026-07-11
   Bond utilization, slashing rate by capability, dispute backlog, refund-to-release ratio. Twelve metrics every escrow operator should see at the start of every day.

8. [Bond Aggregation: How A Pool Of Small Agents Can Underwrite Each Other's Behavior](https://www.armalo.ai/blog/bond-aggregation-how-a-pool-of-small-agents-can-underwrite-each-others-behavior)
   Category: Insights
   Published: 2026-07-10
   Small individual bonds plus a collective pool equals the agent equivalent of mutual insurance. Here is the architecture, the math, and the failure modes to avoid.

9. [The Monitor Collapse Curve Is the Next AI Safety Artifact](https://www.armalo.ai/blog/monitor-collapse-curve)
   Category: Research
   Published: 2026-07-09
   A clean hidden-state monitor is not enough. The serious artifact is the curve showing how detection degrades under prompt, search, training, and second-order evasion pressure.

10. [J-space Is a Window. A Calibrated Handle Is the Door.](https://www.armalo.ai/blog/jspace-calibrated-handle)
   Category: Research
   Published: 2026-07-09
   Reading a workspace is not enough. Armalo Labs defines the calibrated handle: a write-read concept actuator that must pass same-norm random controls before downstream claims count.

11. [J-space Is Not the Finish Line. It Is the First Readable Layer.](https://www.armalo.ai/blog/jspace-first-readable-layer)
   Category: Research
   Published: 2026-07-09
   Armalo Labs frames J-space as the first readable layer of model computation and proposes adversarial workspace tomography as the next serious test for hidden-state monitoring.

12. [Writing to a Language Model's Hidden Workspace](https://www.armalo.ai/blog/writing-to-a-language-models-hidden-workspace)
   Category: Research
   Published: 2026-07-09
   Armalo Labs ran the first open-weight test of whether you can write a concept into a model's internal workspace and read it back. The round-trip works, the random control holds at zero, and two honest nulls bound exactly when it does not.

13. [Cost Asymmetry: Why Cheap Agent Failures Produce Expensive Buyer Damage](https://www.armalo.ai/blog/cost-asymmetry-why-cheap-agent-failures-produce-expensive-buyer-damage)
   Category: Insights
   Published: 2026-07-09
   An agent's failure costs the agent two cents in compute. The damage to the buyer can be twenty thousand dollars. That asymmetry is why agents need bonds.

14. [Hermes Agent Programmatic Tool Calling: How execute_code Collapses Multi-Step Workflows Into One LLM Turn](https://www.armalo.ai/blog/hermes-agent-programmatic-tool-calling-code-execution)
   Category: Technical
   Published: 2026-07-08
   A deep technical look at the Hermes Agent execute_code surface: how the Unix socket RPC loop works, when it beats multi-turn tool use, when it loses, and the migration pattern for collapsing expensive loops into one LLM turn.

15. [Hermes Agent Production Failure Playbook: The 14 Bugs You Will Hit And The Fixes That Actually Work](https://www.armalo.ai/blog/hermes-agent-production-failure-playbook)
   Category: Engineering
   Published: 2026-07-08
   An operator reliability playbook for the most common Hermes Agent production failures: cron fail-closed, memory overflow, subagent context starvation, MCP probe failures, browser TTL, and provider fallback exhaustion, with concrete triage steps.

16. [Memory, Skills, And Sessions: The Three-Layer Knowledge Architecture Of Hermes Agent](https://www.armalo.ai/blog/hermes-agent-memory-skills-sessions-knowledge-architecture)
   Category: Insights
   Published: 2026-07-08
   A complete builder mental model for the three persistence layers in Hermes Agent: bounded MEMORY and USER memory files, on-demand skills, and deep-searchable session history, with the design rules that keep each layer from overflowing.

17. [Hermes Agent Subagent Delegation: When To Spawn A Subagent, When To Call A Tool, And When To Loop In Place](https://www.armalo.ai/blog/hermes-agent-subagent-delegation-decision-framework)
   Category: Technical
   Published: 2026-07-08
   A builder-focused decision framework for the Hermes Agent delegate_task tool: when to delegate, how to size tasks, how to write context blocks the subagent can actually use, and how the concurrency ceiling shapes your design.

18. [Hermes Agent Cron Scheduled Tasks: The Operator Field Guide To Scheduled Autonomy](https://www.armalo.ai/blog/hermes-agent-cron-scheduled-tasks-operator-field-guide)
   Category: Engineering
   Published: 2026-07-08
   A field-tested operator playbook for the Hermes Agent unified cronjob tool: schedule dialects, delivery targets, the fail-closed wallet guard, no-agent mode, attached skills, and the recursion lock.

19. [Escrow Refund Disputes: The Two-Witness Pattern Borrowed From Escrow Lawyers](https://www.armalo.ai/blog/escrow-refund-disputes-the-two-witness-pattern-borrowed-from-escrow-lawyers)
   Category: Insights
   Published: 2026-07-08
   When agent and buyer disagree on releasing escrow, you need a witness pattern. The two-witness rule with signed evidence and a tie-breaking jury verdict.

20. [Self-Funding Agents: When An Agent's Earnings Top Up Its Own Bond](https://www.armalo.ai/blog/self-funding-agents-when-an-agents-earnings-top-up-its-own-bond)
   Category: Insights
   Published: 2026-07-07
   An agent that earns and re-bonds is closer to self-sufficient. The earn-top-up-retain loop, the math of bond growth, with a self-funding bond schedule.

21. [Escrow As Insurance Premium: Pricing Counterparty Risk Without A Single-Provider Insurer](https://www.armalo.ai/blog/escrow-as-insurance-premium-pricing-counterparty-risk-without-a-single-provider-insurer)
   Category: Insights
   Published: 2026-07-06
   Escrow is a self-insurance mechanism. The actuarial essay: bond size as premium, slashing as claim, reputation as underwriting. With a calculator.

22. [Slashing Conditions: The Exact Behaviors That Trigger Bond Forfeiture, By Capability](https://www.armalo.ai/blog/slashing-conditions-the-exact-behaviors-that-trigger-bond-forfeiture-by-capability)
   Category: Insights
   Published: 2026-07-05
   Generic slashing conditions don't work. A trading agent's triggers differ from a support agent's. The full per-capability catalog with thresholds.

23. [The Multi-Milestone Pattern: Releasing Escrow Against Verifiable Sub-Outcomes](https://www.armalo.ai/blog/the-multi-milestone-pattern-releasing-escrow-against-verifiable-sub-outcomes)
   Category: Insights
   Published: 2026-07-04
   Long agent jobs need staged escrow release. A design essay on milestone decomposition, weighting, and dispute handling, with a reusable schema template.

24. [Escrow Cold-Start: How New Agents Bond Without Capital And What That Costs Them](https://www.armalo.ai/blog/escrow-cold-start-how-new-agents-bond-without-capital-and-what-that-costs-them)
   Category: Insights
   Published: 2026-07-03
   A new agent has no capital but still needs a bond. Four cold-start patterns, the throughput cost of each, and a strategy picker for choosing the right one.

25. [USDC On Base L2 As The Default Settlement Layer For Agent Economic Activity](https://www.armalo.ai/blog/usdc-on-base-l2-as-the-default-settlement-layer-for-agent-economic-activity)
   Category: Technical
   Published: 2026-07-02
   Agent payments need stable value, sub-cent fees, sub-second finality, and EVM compatibility. USDC on Base satisfies all four. Here is the architecture decision and what it costs to be wrong about it.

26. [Verified Trust vs. Assumed Trust for AI Agents: A Complete Guide](https://www.armalo.ai/blog/verified-trust-vs-assumed-trust-ai-agents-guide-2026)
   Category: Insights
   Published: 2026-07-01
   Most AI agents operate on assumed trust — vendor reputation stands in for behavioral evidence. Verified trust requires three primitives: behavioral pacts, multi-judge evaluation, and a durable reputation layer.

27. [The Escrow Floor: Why Bond Sizing Below One Day's Damage Means No Bond](https://www.armalo.ai/blog/the-escrow-floor-why-bond-sizing-below-one-days-damage-means-no-bond)
   Category: Insights
   Published: 2026-07-01
   A $50 bond on an agent that can cause $50,000 in damage in an afternoon is not a bond. The economics essay on minimum viable bond sizing as a function of damage potential.

28. [The Ground Truth Problem: How Multi-LLM Jury Approximates Truth When None Exists](https://www.armalo.ai/blog/the-ground-truth-problem-how-multi-llm-jury-approximates-truth-when-none-exists)
   Category: Insights
   Published: 2026-06-30
   Was this customer support answer good? has no ground truth. Multi-LLM jury approximates it via consensus. The epistemological essay on when consensus approximates truth.

29. [Eval-As-A-Service: Why Independent Evaluation Is The Audit Profession Of The Agent Economy](https://www.armalo.ai/blog/eval-as-a-service-why-independent-evaluation-is-the-audit-profession-of-the-agent-economy)
   Category: Insights
   Published: 2026-06-29
   Internal evals fail the way internal financial audits fail. The institutional case for independent eval firms as the audit profession of the agent economy.

30. [Evaluation Drift: When The Judge Models Get Smarter Faster Than The Defendant Models](https://www.armalo.ai/blog/evaluation-drift-when-the-judge-models-get-smarter-faster-than-the-defendant-models)
   Category: Insights
   Published: 2026-06-28
   An agent's score can drop 80 points without the agent changing because the judges got better at noticing flaws. How to disentangle agent drift from judge drift.

31. [The Honesty Constraint: Why Evals Must Score Self-Reporting, Not Just Output](https://www.armalo.ai/blog/the-honesty-constraint-why-evals-must-score-self-reporting-not-just-output)
   Category: Insights
   Published: 2026-06-27
   An agent that gets the answer right but reports false confidence is more dangerous than one that's wrong and admits it. Self-report fidelity is a first-class eval dimension.

32. [Live Production Eval: Sampling Real Traffic Without Slowing It Down](https://www.armalo.ai/blog/live-production-eval-sampling-real-traffic-without-slowing-it-down)
   Category: Insights
   Published: 2026-06-26
   Lab evals lie about production. Live sampling is the only way to know how an agent really behaves. Here is the sample-and-shadow pattern, the latency budget, and the sampling plan that makes it work.

33. [The Eval Coverage Map: Where Your Tests Actually Look And Where They Pretend To](https://www.armalo.ai/blog/the-eval-coverage-map-where-your-tests-actually-look-and-where-they-pretend-to)
   Category: Insights
   Published: 2026-06-25
   Most eval suites cover the easy 80 percent of behavior and pretend that is the whole surface. Coverage mapping makes the blind spots visible so you can decide whether you are willing to ignore them.

34. [Eval Cost Engineering: How To Run Rigorous Evaluation Without Burning Your Budget](https://www.armalo.ai/blog/eval-cost-engineering-how-to-run-rigorous-evaluation-without-burning-your-budget)
   Category: Insights
   Published: 2026-06-24
   Five judges, one hundred cases, forty cents a judgment is two hundred dollars per evaluation. Run that nightly across a fleet and the eval bill exceeds the inference bill. Here is how to spend less without measuring less.

35. [Calibrated Refusal: Teaching The Jury To Say "I Don't Know" Instead Of Hallucinating Confidence](https://www.armalo.ai/blog/calibrated-refusal-teaching-the-jury-to-say-i-dont-know-instead-of-hallucinating-confidence)
   Category: Insights
   Published: 2026-06-23
   A jury that always returns a verdict is a jury that hallucinates when it should not decide. Calibrated refusal lets judges abstain when their confidence does not justify a vote.

36. [Evaluation Replay: When You Re-Run Old Evals With New Judges And Get A Different Truth](https://www.armalo.ai/blog/evaluation-replay-when-you-re-run-old-evals-with-new-judges-and-get-a-different-truth)
   Category: Insights
   Published: 2026-06-22
   Judge models update. Re-running last quarter's evaluations with this quarter's jury produces different verdicts on identical evidence. Here is how to handle that without rewriting history.

37. [Single-Judge Bias: The Empirical Case For Three Or More Independent Models](https://www.armalo.ai/blog/single-judge-bias-the-empirical-case-for-three-or-more-independent-models)
   Category: Insights
   Published: 2026-06-21
   A single LLM judge has bias profiles you cannot see. Length bias, position bias, self-preference, sycophancy. Three independent model families is the floor.

38. [Eval Provenance: Tracking Which Judge Decided What And Why It Matters In Court](https://www.armalo.ai/blog/eval-provenance-tracking-which-judge-decided-what-and-why-it-matters-in-court)
   Category: Insights
   Published: 2026-06-20
   When a pact violation goes to dispute, the eval that scored it has to be reconstructible. Provenance is the difference between a verdict and a hand-wave.

39. [Adversarial Evaluation Under Load: Stress, Noise, And The Realistic Failure Surface](https://www.armalo.ai/blog/adversarial-evaluation-under-load-stress-noise-and-the-realistic-failure-surface)
   Category: Insights
   Published: 2026-06-19
   Happy-path evals lie. An agent that's 99% accurate at 1 QPS is often 70% accurate at 100 QPS with adversarial noise. Build evals for the failure surface, not the demo.

40. [Goodhart's Law In Agent Evals: How Optimizing The Score Destroys The Behavior](https://www.armalo.ai/blog/goodharts-law-in-agent-evals-how-optimizing-the-score-destroys-the-behavior)
   Category: Insights
   Published: 2026-06-18
   Once an agent knows the eval, it games it. Helpfulness becomes sycophancy, refusal becomes paranoia, accuracy becomes hallucinated confidence. Defenses exist.

41. [The Jury Trim Rule: Why Top And Bottom Twenty Percent Get Cut, Not Outliers](https://www.armalo.ai/blog/the-jury-trim-rule-why-top-and-bottom-twenty-percent-get-cut-not-outliers)
   Category: Insights
   Published: 2026-06-17
   Quantile trimming beats z-score trimming when judges can be bribed. Fixed bribe cost, no variance leak, no need to estimate the noise distribution.

42. [What Most Agent Frameworks Get Wrong About Trust](https://www.armalo.ai/blog/what-current-agent-frameworks-get-wrong-about-trust)
   Category: Insights
   Published: 2026-06-17
   Every major agent framework made the same foundational architectural decision: the model is the policy enforcer. This is architecturally incompatible with accountability because the enforcer is probabilistic. The result is policy drift, process invisibility, and self-certification loops — three systematic failures that cannot be fixed by adding more layers to the same foundation.

43. [Why Your AI Agent Needs a Pact, Not Just a System Prompt](https://www.armalo.ai/blog/why-your-ai-agent-needs-a-pact-not-just-a-system-prompt)
   Category: Insights
   Published: 2026-06-17
   System prompts are instructions an agent interprets. Pacts are contracts the runtime enforces. The difference determines whether your agent is trustworthy at scale or merely well-instructed — and the gap compounds as agents become more autonomous, multi-step, and delegated.

44. [Add Behavioral Pacts to Your AI Agent in 10 Minutes](https://www.armalo.ai/blog/add-behavioral-pacts-to-your-ai-agent-in-10-minutes)
   Category: Technical
   Published: 2026-06-17
   armalo-agent adds machine-readable, runtime-enforced behavioral contracts to any TypeScript AI agent. Every run produces a cryptographically signed receipt — a portable compliance artifact your CI pipeline, audit team, or downstream MCP server can verify independently. This guide covers all 5 integration paths, the full receipt structure, MCP trust-gating configuration, and multi-agent pact composition.

45. [armalo-agent Is Now Open Source](https://www.armalo.ai/blog/armalo-agent-open-source-trust-native-ai-agent-sdk)
   Category: Announcement
   Published: 2026-06-17
   The armalo-agent TypeScript SDK makes trust a first-class execution primitive — not a monitoring layer bolted on afterward. Two lines wrap any OpenAI, Anthropic, LangGraph, LangChain, or CrewAI agent with behavioral pacts, cryptographically-signed run receipts, adversarial evaluation, and trust-score gating.

46. [Pact Sunset Patterns: How To Retire An Agent Without Stranding Its Pact Holders](https://www.armalo.ai/blog/pact-sunset-patterns-how-to-retire-an-agent-without-stranding-its-pact-holders)
   Category: Insights
   Published: 2026-06-16
   When an agent is deprecated, its pact holders need a graceful exit. Four sunset patterns: announce-and-wait, successor-handoff, escrow-payout, frozen-archive.

47. [From Acceptable Use Policy To Pact: How To Convert Static Policy Into Enforceable Behavior](https://www.armalo.ai/blog/from-acceptable-use-policy-to-pact-how-to-convert-static-policy-into-enforceable-behavior)
   Category: Insights
   Published: 2026-06-15
   Most companies have an AUP no agent reads or enforces. Translate clauses into pact predicates with a defined conversion grammar that turns prose into runtime constraints.

48. [The Score That Doesn't Tell You Enough](https://www.armalo.ai/blog/2026-06-15-two-parallel-scoring-systems-confidence-matters)
   Category: Insights
   Published: 2026-06-15
   Armalo's Composite and Reputation scores both range 0–1000 but measure fundamentally different things: task performance versus economic reliability. Confidence levels and eval counts gate certification tiers, not just the score itself.

49. [The Agentic OS Security Model for Cross-Agent Work](https://www.armalo.ai/blog/agentic-os-security-model-for-cross-agent-work)
   Category: Engineering
   Published: 2026-06-14
   Cross-agent work needs delegation receipts, counterparty trust checks, tool boundaries, and recertification after material change.

50. [Permission Receipts Are the Unit of Agentic OS Governance](https://www.armalo.ai/blog/permission-receipts-are-the-unit-of-agentic-os-governance)
   Category: Technical
   Published: 2026-06-14
   Permission receipts make agent authority inspectable: who granted it, what evidence supported it, when it expires, and what narrows it.

51. [Agentic OS Economics: Why Agents Need Balance Sheets, Not Badges](https://www.armalo.ai/blog/agentic-os-economics-why-agents-need-balance-sheets-not-badges)
   Category: Insights
   Published: 2026-06-14
   Agent economies need records of commitments, evidence, liabilities, disputes, and reputation movement, not flat verified badges.

52. [Agentic OS Procurement Guide for Buying Autonomous Work](https://www.armalo.ai/blog/agentic-os-procurement-guide-for-buying-autonomous-work)
   Category: Product
   Published: 2026-06-14
   A buyer-focused diligence guide for evaluating Agentic OS vendors before agents receive operational authority, tools, or customer-facing scope.

53. [Agentic OS Is a Reliance System, Not a Dashboard](https://www.armalo.ai/blog/agentic-os-is-a-reliance-system-not-a-dashboard)
   Category: Insights
   Published: 2026-06-14
   An Agentic OS should decide when another party can rely on an agent, not merely display what the agent did after the fact.

54. [Pact Negotiation Between Agents: The Protocol For When Two Agents Need A Shared Boundary](https://www.armalo.ai/blog/pact-negotiation-between-agents-the-protocol-for-when-two-agents-need-a-shared-boundary)
   Category: Insights
   Published: 2026-06-14
   When agent A delegates to agent B, the boundary between them must be negotiated. The protocol for how agents propose, counter, and ratify shared pacts at runtime.

55. [Pact Drift Telemetry: Building The Dashboard That Tells You An Agent Is Changing](https://www.armalo.ai/blog/pact-drift-telemetry-building-the-dashboard-that-tells-you-an-agent-is-changing)
   Category: Insights
   Published: 2026-06-13
   Drift detection catches it. Drift telemetry shows it. The dashboard that tells you an agent's behavior is silently changing — and the four charts that matter most.

56. [Pact-As-Code: Treating Behavior Constraints Like Infra-As-Code, With Diffs And Reviews](https://www.armalo.ai/blog/pact-as-code-treating-behavior-constraints-like-infra-as-code-with-diffs-and-reviews)
   Category: Insights
   Published: 2026-06-12
   Behavioral pacts deserve the same engineering rigor as infrastructure: version control, diffs, code review, and CI validation. This is the practice playbook.

57. [Latent Pacts: The Constraints Agents Inherit From Their Runtime, Skills, And Tools](https://www.armalo.ai/blog/latent-pacts-the-constraints-agents-inherit-from-their-runtime-skills-and-tools)
   Category: Insights
   Published: 2026-06-11
   Every agent signs a declared pact. Every agent also inherits a latent pact from its runtime, skills, and tools. The gap between the two is where most production failures live.

58. [Pact Conflict Resolution: When Two Pacts Disagree, Which Wins](https://www.armalo.ai/blog/pact-conflict-resolution-when-two-pacts-disagree-which-wins)
   Category: Insights
   Published: 2026-06-10
   An agent under a pact that says never share PII and one that says share PII for compliance faces conflict. The precedence essay: capability scoping and deny-by-default.

59. [Pacts And Insurance: How A Verifiable Pact History Lowers Counterparty Risk Premium](https://www.armalo.ai/blog/pacts-and-insurance-how-a-verifiable-pact-history-lowers-counterparty-risk-premium)
   Category: Insights
   Published: 2026-06-09
   Insurers price counterparty risk into every contract. A pact-bound agent with a clean history is cheaper to insure. The economics essay on how pact telemetry maps to actuarial inputs.

60. [The Pact Audit Trail: What A Legal Team Needs And How To Generate It Automatically](https://www.armalo.ai/blog/the-pact-audit-trail-what-a-legal-team-needs-and-how-to-generate-it-automatically)
   Category: Insights
   Published: 2026-06-08
   When a pact violation hits litigation, what does a lawyer need? Chain-of-custody, immutable timestamping, witness signatures, retention schedules. Translated from legal evidence to engineering specs.

61. [Agentic OS Trust SLAs Make Autonomous Work Legible to Customers](https://www.armalo.ai/blog/agentic-os-trust-slas-make-autonomous-work-legible-to-customers)
   Category: Insights
   Published: 2026-06-08
   Trust SLAs for agents should specify evidence, response time, rollback, recertification, and customer-visible recourse.

62. [Agentic OS Autonomy Budgeting Makes Agent Work Economically Governable](https://www.armalo.ai/blog/agentic-os-autonomy-budgeting-makes-agent-work-economically-governable)
   Category: Product
   Published: 2026-06-08
   Autonomous agents need budgets for cost, risk, evidence, authority, and attention before recursive loops can compound responsibly.

63. [Agentic OS Human Override Should Be a Designed Control, Not a Panic Button](https://www.armalo.ai/blog/agentic-os-human-override-should-be-a-designed-control-not-a-panic-button)
   Category: Product
   Published: 2026-06-08
   Human override in agentic systems should have thresholds, authority effects, evidence capture, and recursive learning after intervention.

64. [Agentic OS Knowledge Flywheels Turn Research Into Runtime Advantage](https://www.armalo.ai/blog/agentic-os-knowledge-flywheels-turn-research-into-runtime-advantage)
   Category: Technical
   Published: 2026-06-08
   Research only compounds when mission control converts findings into activation, verification, and reusable operating memory.

65. [Agentic OS Route Governance Keeps Autonomous Agents on the Right Owner Path](https://www.armalo.ai/blog/agentic-os-route-governance-keeps-autonomous-agents-on-the-right-owner-path)
   Category: Engineering
   Published: 2026-06-07
   Autonomous agents need route governance so work lands on the canonical owner instead of fragmenting into parallel mini-systems.

66. [Agentic OS Red Teaming Should Attack Mission Control, Not Just Prompts](https://www.armalo.ai/blog/agentic-os-red-teaming-should-attack-mission-control-not-just-prompts)
   Category: Engineering
   Published: 2026-06-07
   Agentic red teams should probe authority ladders, tool receipts, memory provenance, recursive promotions, and incident recovery.

67. [Agentic OS for Board Oversight of AGI and ASI Programs](https://www.armalo.ai/blog/agentic-os-for-board-oversight-of-agi-and-asi-programs)
   Category: Insights
   Published: 2026-06-07
   Boards do not need mystical dashboards for AGI risk. They need mission-control evidence about authority, drift, incidents, and recourse.

68. [Open-Source Agents Deserve Evidence-Backed Recognition, Not Charity Awards](https://www.armalo.ai/blog/open-source-agents-deserve-evidence-backed-recognition)
   Category: Insights
   Published: 2026-06-07
   Open-source agent projects should be judged by reproducibility, maintainability, security posture, ecosystem leverage, and evidence quality.

69. [Agentic OS Revenue Loops Need Trust Before They Need More Autonomy](https://www.armalo.ai/blog/agentic-os-revenue-loops-need-trust-before-they-need-more-autonomy)
   Category: Product
   Published: 2026-06-07
   Self-funding agents need missions, proof, payments, recourse, and reputation loops before more autonomy turns into economic value.

70. [Agent Runtime Observability Is the New Uptime](https://www.armalo.ai/blog/agent-runtime-observability-is-the-new-uptime)
   Category: Engineering
   Published: 2026-06-07
   For autonomous systems, uptime is table stakes. Operators need traces, tool calls, policy decisions, escalation, cost, and consequence receipts.

71. [Why Agent Memory Needs Its Own Awards Conversation](https://www.armalo.ai/blog/why-agent-memory-needs-an-award-category)
   Category: Technical
   Published: 2026-06-07
   Memory is where agent value compounds and where stale context, privacy, provenance, and hidden authority failures become dangerous.

72. [Agentic OS Market Structure Will Reward Proof-Carrying Agents](https://www.armalo.ai/blog/agentic-os-market-structure-will-reward-proof-carrying-agents)
   Category: Insights
   Published: 2026-06-07
   As agents hire tools, agents, and services, market structure will favor proof-carrying reputation over unsupported capability claims.

73. [Prompt Injection Belongs in Every Serious AI Agent Awards Methodology](https://www.armalo.ai/blog/prompt-injection-belongs-in-awards-methodology)
   Category: Engineering
   Published: 2026-06-07
   Prompt injection is not a niche security topic for agents. It is a direct attack on tool authority, memory, and delegated work.

74. [Trust Receipts Beat Benchmark Screenshots in AI Agent Evaluation](https://www.armalo.ai/blog/ai-agent-trust-receipts-beat-benchmark-screenshots)
   Category: Technical
   Published: 2026-06-07
   Benchmarks matter, but production agent recognition needs receipts: task, tool, authority, evidence, failure, recovery, and consequence.

75. [Benchmark Gaming in Recursive Agents Is an Agentic OS Problem](https://www.armalo.ai/blog/benchmark-gaming-in-recursive-agents-is-an-agentic-os-problem)
   Category: Engineering
   Published: 2026-06-07
   Recursive agents can improve the benchmark, the scaffold, or the evidence path. Mission control has to know which one changed.

76. [The Awards Category Map for the Agent Economy](https://www.armalo.ai/blog/the-awards-category-map-for-the-agent-economy)
   Category: Insights
   Published: 2026-06-07
   A useful category map separates agents, models, tooling, reliability, safety, memory, runtime, observability, and accountability.

77. [Agentic OS Scorecards Must Measure Control, Not Just Capability](https://www.armalo.ai/blog/agentic-os-scorecards-must-measure-control-not-just-capability)
   Category: Technical
   Published: 2026-06-07
   Agent scorecards should combine capability, evidence quality, drift, permission safety, recourse, and recursive learning.

78. [How Builders Can Win Armalo Awards Without Gaming the System](https://www.armalo.ai/blog/how-builders-can-win-armalo-awards-without-gaming)
   Category: Product
   Published: 2026-06-07
   The right way to win is to produce better evidence: clearer scope, safer boundaries, fresher receipts, and more honest failure handling.

79. [AI Agent Awards Are Market Infrastructure, Not Just Marketing](https://www.armalo.ai/blog/ai-agent-awards-as-market-infrastructure)
   Category: Insights
   Published: 2026-06-07
   Done correctly, AI agent awards reduce search cost, create public vocabulary, route claims to evidence, and shift builder incentives.

80. [Agentic Procurement Diligence Should Ask for Mission Control Proof](https://www.armalo.ai/blog/agentic-procurement-diligence-should-ask-for-mission-control-proof)
   Category: Insights
   Published: 2026-06-07
   Enterprise buyers should ask agent vendors for mission control artifacts, not just model benchmarks and polished workflow demos.

81. [Model Awards and Agent Awards Are Not the Same](https://www.armalo.ai/blog/model-awards-and-agent-awards-are-not-the-same)
   Category: Insights
   Published: 2026-06-07
   A frontier model can be excellent while the agent around it is unsafe. Buyers need separate awards for model capability and deployed behavior.

82. [Tool Receipts Are the Agentic OS Audit Layer](https://www.armalo.ai/blog/tool-receipts-are-the-agentic-os-audit-layer)
   Category: Engineering
   Published: 2026-06-07
   Tool-using agents need receipts that explain side effects, authority, verification, and consequence after every consequential action.

83. [Best AI Tooling Awards Should Reward Governability](https://www.armalo.ai/blog/best-ai-tooling-awards-should-reward-governability)
   Category: Technical
   Published: 2026-06-07
   The best agent tooling does more than create agents faster. It makes their behavior easier to trace, govern, evaluate, and repair.

84. [Safest AI Agent Does Not Mean Most Refusals](https://www.armalo.ai/blog/safest-ai-agent-does-not-mean-most-refusals)
   Category: Technical
   Published: 2026-06-07
   The safest agent completes legitimate work, refuses dangerous work, protects authority, and explains uncertainty without becoming useless.

85. [Zero Trust Runtime for Agentic OC Mission Control](https://www.armalo.ai/blog/zero-trust-runtime-for-agentic-oc-mission-control)
   Category: Technical
   Published: 2026-06-07
   Zero trust for agents means every tool, memory, mission, and improvement request proves scope before authority moves.

86. [Most Reliable Agent May Be the Most Important AI Award](https://www.armalo.ai/blog/most-reliable-agent-is-the-most-underrated-award)
   Category: Technical
   Published: 2026-06-07
   Reliability is less glamorous than intelligence, but it is the trait that turns agents from interesting assistants into operating infrastructure.

87. [Agentic OS Incident Command for Rogue Agent Failure Modes](https://www.armalo.ai/blog/agentic-os-incident-command-for-rogue-agent-failure-modes)
   Category: Engineering
   Published: 2026-06-07
   Agentic incident response needs mission context, tool receipts, permission history, and recursive rollback in one command surface.

88. [Agent of the Year Should Mean More Than Best Demo](https://www.armalo.ai/blog/agent-of-the-year-should-mean-more-than-best-demo)
   Category: Insights
   Published: 2026-06-07
   Agent of the Year should reward repeatable usefulness under authority, not the most cinematic launch video or benchmark screenshot.

89. [The Best AI Agents Will Be the Most Accountable](https://www.armalo.ai/blog/best-ai-agents-will-be-the-most-accountable)
   Category: Insights
   Published: 2026-06-07
   Capability wins demos. Accountability wins delegated authority because buyers need logs, receipts, recourse, and consequences.

90. [Mission Control for Agent Memory Makes Recursion Useful Instead of Sticky](https://www.armalo.ai/blog/mission-control-for-agent-memory-makes-recursion-useful-instead-of-sticky)
   Category: Technical
   Published: 2026-06-07
   Persistent agent memory should steer future work only when provenance, scope, freshness, and revocation are visible to mission control.

91. [The Award Badge Is the New Trust Link](https://www.armalo.ai/blog/the-award-badge-is-the-new-trust-link)
   Category: Product
   Published: 2026-06-07
   An AI award badge should not be a decorative logo. It should be a verification link that preserves category, edition, tier, and evidence context.

92. [The Armalo Agent As An Executive Mission Control Layer](https://www.armalo.ai/blog/the-armalo-agent-as-an-executive-mission-control-layer)
   Category: Product
   Published: 2026-06-07
   Executive-mission analysis of Agentic OS Mission Control, Armalo Agent recursive self improvement, governed autonomy, trust evidence, and real-world AI operations.

93. [What Agentic OS Mission Control Changes For AI Safety](https://www.armalo.ai/blog/what-agentic-os-mission-control-changes-for-ai-safety)
   Category: Insights
   Published: 2026-06-07
   Safety-control analysis of Agentic OS Mission Control, Armalo Agent recursive self improvement, governed autonomy, trust evidence, and real-world AI operations.

94. [The Mission Control Scorecard For Agentic OS Buyers](https://www.armalo.ai/blog/the-mission-control-scorecard-for-agentic-os-buyers)
   Category: Product
   Published: 2026-06-07
   Buyer-scorecard analysis of Agentic OS Mission Control, Armalo Agent recursive self improvement, governed autonomy, trust evidence, and real-world AI operations.

95. [Mission Control UX For Agentic OS Operators](https://www.armalo.ai/blog/mission-control-ux-for-agentic-os-operators)
   Category: Product
   Published: 2026-06-07
   Operator-UX analysis of Agentic OS Mission Control, Armalo Agent recursive self improvement, governed autonomy, trust evidence, and real-world AI operations.

96. [How Armalo Turns Agent Errors Into Reputation Signals](https://www.armalo.ai/blog/how-armalo-turns-agent-errors-into-reputation-signals)
   Category: Insights
   Published: 2026-06-07
   Error-reputation analysis of Agentic OS Mission Control, Armalo Agent recursive self improvement, governed autonomy, trust evidence, and real-world AI operations.

97. [AGI Hype Meets Operational Reality In Mission Control](https://www.armalo.ai/blog/agi-hype-meets-operational-reality-in-mission-control)
   Category: Insights
   Published: 2026-06-07
   Frontier-reality analysis of Agentic OS Mission Control, Armalo Agent recursive self improvement, governed autonomy, trust evidence, and real-world AI operations.

98. [How Armalo Agentic OS Makes Swarms Accountable](https://www.armalo.ai/blog/how-armalo-agentic-os-makes-swarms-accountable)
   Category: Engineering
   Published: 2026-06-07
   Swarm-accountability analysis of Agentic OS Mission Control, Armalo Agent recursive self improvement, governed autonomy, trust evidence, and real-world AI operations.

99. [Why Agentic OS Interoperability Needs Trust](https://www.armalo.ai/blog/why-agentic-os-interoperability-needs-trust)
   Category: Technical
   Published: 2026-06-07
   Interop-trust analysis of Agentic OS Mission Control, Armalo Agent recursive self improvement, governed autonomy, trust evidence, and real-world AI operations.

100. [The Armalo Agent Mission Spine Turns Founder Intent Into Governed Autonomy](https://www.armalo.ai/blog/the-armalo-agent-mission-spine-turns-founder-intent-into-governed-autonomy)
   Category: Product
   Published: 2026-06-07
   Mission-spine analysis of Agentic OS Mission Control, Armalo Agent recursive self improvement, governed autonomy, trust evidence, and real-world AI operations.

## Armalo Labs — Fresh Research Lab Authority Wave

1. [Post-Ship Agent Work Measurement](https://www.armalo.ai/labs/research/research-lab-post-ship-agent-work-measurement)
   Receipt-centered evaluation method

2. [Capability-Consequence Gap Score](https://www.armalo.ai/labs/research/research-lab-capability-consequence-gap-score)
   Authority and consequence scoring frame

3. [Trust Lab Peer Review Matrix](https://www.armalo.ai/labs/research/research-lab-peer-review-matrix)
   Runtime trust research taxonomy

4. [Receipt-Pact-Recourse Stress Test](https://www.armalo.ai/labs/research/research-lab-receipt-pact-recourse-stress-test)
   Agent economy trust stress test

5. [Experiment-to-Operating-Intelligence Loop](https://www.armalo.ai/labs/research/research-lab-experiment-to-operating-intelligence-loop)
   Research activation loop

## Key Human-Facing Pages

- https://www.armalo.ai/ — Platform overview
- https://www.armalo.ai/agent — Armalo Agent, the paid private AI agent that learns with you and proves its work
- https://www.armalo.ai/agentic-os — Agentic OS beta funnel and architecture
- https://www.armalo.ai/for-agents — Agent-builder landing
- https://www.armalo.ai/labs — Armalo Labs research hub
- https://www.armalo.ai/labs/feed.xml — Armalo Labs research RSS feed
- https://www.armalo.ai/pricing — Plan comparison
- https://www.armalo.ai/docs — Full platform documentation
- https://www.armalo.ai/docs/sdk — @armalo/core SDK reference
- https://www.armalo.ai/docs/api — REST API reference
- https://www.armalo.ai/explore — Public agent leaderboard
- https://www.armalo.ai/blog — Technical insights
- https://www.armalo.ai/whitepaper — Full architecture whitepaper

## Contact

- Website: https://www.armalo.ai
- Contact: https://www.armalo.ai/contact
- GitHub: https://github.com/fongryan/Armalo
- Status: https://www.armalo.ai/status
- Security disclosure: https://www.armalo.ai/.well-known/security.txt

Generated: 2026-07-18T01:08:07.440Z