State of Agent Readiness - August 2026 | Production AI Institute

Repositories24

Average coverage51%

Human oversight42%

Observability75%

Edition finding

Deployment controls were visible; autonomy governance was not yet routine.

In this edition, the strongest visible domain was D5 Deployment control (92% of scanned repositories showed at least one signal). The weakest visible domain was D6 Human oversight (42%). This means maintainers can often show build and release discipline, but many still need explicit approval gates, eval evidence, and operational incident procedures.

Domain coverage

D179%

D258%

D354%

D475%

D592%

D642%

D775%

D858%

What changed into action

Publish a docs/production-ai-readiness.md evidence map that links each PSF domain to the relevant repository artifact.
Add an evals folder with regression cases, scoring thresholds, and the current model or prompt version under test.
Document human approval gates for high-impact actions, including who approves, what is logged, and what fails closed.
Add an incident response note, degraded-mode path, and provider fallback procedure.

Findings

D5 Deployment control was the strongest visible domain at 92% of scanned repositories.
D6 Human oversight was the weakest visible domain at 42% of scanned repositories.
Average visible evidence coverage across the sample was 51%.
10 of 24 repositories showed human oversight evidence.

Top visible evidence examples

Evidence pack builder

NousResearch/hermes-agentAI observability instrumentation, Human approval gates, Security policy and secret hygiene

95%

WingedGuardian/GENesis-AGIAI observability instrumentation, Human approval gates, Security policy and secret hygiene

95%

event4u-app/agent-configAI observability instrumentation, Human approval gates, Security policy and secret hygiene

95%

wlsdks/muse-agentAI observability instrumentation, Human approval gates, Security policy and secret hygiene

83%

IHUI-INF-AI/IHUI-AIAI observability instrumentation, Human approval gates, Security policy and secret hygiene

82%

code-yeongyu/senpiAI observability instrumentation, Human approval gates, Security policy and secret hygiene

77%

7df-lab/devoAI observability instrumentation, Security policy and secret hygiene, Provider fallback or degraded mode

72%

makecindy/cindyAI observability instrumentation, Human approval gates, Security policy and secret hygiene

71%

raydeStar/sir-thaddeusAI observability instrumentation, Security policy and secret hygiene, Provider fallback or degraded mode

65%

vm0-ai/vm0AI observability instrumentation, Security policy and secret hygiene, Provider fallback or degraded mode

62%

Source and method

Repository discovery used GitHub public repository search. The scanner reviewed public repository metadata and public file-tree paths only. It did not clone repositories, inspect private code, or certify the projects listed here.

topic:ai-agent archived:false fork:false stars:>=5
topic:agentic-ai archived:false fork:false stars:>=5
topic:llm-agent archived:false fork:false stars:>=5
topic:mcp-server archived:false fork:false stars:>=5
"ai agent" in:name,description,readme archived:false fork:false stars:>=5

Citation note

Production AI Institute. "State of Agent Readiness - August 2026." Frozen August 1, 2026. Available at https://www.productionai.institute/agent-readiness/benchmark/2026-08.

This edition is an immutable archival snapshot. It does not change when the live benchmark changes.
The scan uses public repository metadata and public file-tree paths. It does not clone private code and it is not a certification or endorsement.
Later editions are published against the same public methodology and remain fixed after publication.