Experiment with CLI usage heuristics

[scottopell]

Summary

Workflow Automation • View in Workflow Automation

Add an Agent CLI LLM-user heuristic flow and route it into cross-org Agent Telemetry, with scoring now explicitly prioritizing temporal behavior over all other signals.

Changes

• Added a new CLI heuristic engine at pkg/cli/heuristic/llm_user_heuristic.go.
  • Computes an LLM-user heuristic score for CLI diagnostics sessions.
  • Uses session state persisted in /tmp/dd-agent-cli-heuristic-state.json.
  • Tracks temporal signals and execution/syntax signals.
• Added heuristic tests at pkg/cli/heuristic/llm_user_heuristic_test.go.
• Added a new internal Agent CMD endpoint at /agent/cli-heuristic in comp/api/commonendpoints/impl/common_endpoints.go.
  • Endpoint forwards payload to Agent Telemetry via SendEvent("llm_cli_heuristic", payload).
• Registered a new default agent telemetry event in comp/core/agenttelemetry/impl/config.go:
  • name: llm_cli_heuristic
  • request_type: agent-cli-heuristic
  • payload_key: agent_cli_heuristic
• Integrated CLI emission points:
  • cmd/agent/subcommands/status/command.go posts heuristic payload for agent status.
  • pkg/cli/subcommands/health/command.go posts heuristic payload for agent health.
• Added endpoint test coverage in comp/api/commonendpoints/impl/common_endpoints_test.go.

Follow-up tuning in this update

• Removed environment-based signals from scoring:
  • TERM
  • NO_COLOR / inferred color support
  • terminal size (COLUMNS/LINES)
• Kept syntax/execution signals:
  • machine-first flags on first invocation (--json, -j, -p, --pretty-json)
  • parent-process pipe filters to jq/grep/yq
• Reweighted scoring to make temporal behavior dominant:
  • very fast inter-command latency for complex diagnostics: +60
  • very low shell aging: +30
  • machine-first flags: +6
  • pipe filters: +4

Testing

• go test -tags test ./pkg/cli/heuristic
• go test -tags test ./pkg/cli/subcommands/health ./cmd/agent/subcommands/status ./comp/api/commonendpoints/impl ./comp/core/agenttelemetry/impl
• Format tool run (goimports/gofmt).
• Previous lint limitation remains environment-related (golangci-lint/go toolchain mismatch), but targeted builds and tests pass.

---

PR by Bits - View session in Datadog

Comment @DataDog to request changes

[datadog-datadog-prod-us1-2]

View session in Datadog

Bits Dev status: ✅ Done

CI Auto-fix: Disabled | Enable

Comment @DataDog to request changes

[datadog-prod-us1-4]

I can only run on private repositories.

[github-actions]

Thank you for your submission, we really appreciate it. Like many open-source projects, we ask that you sign our Contributor License Agreement before we can accept your contribution. You can sign the CLA by just posting a Pull Request Comment same as the below format.

---

I have read the CLA Document and I hereby sign the CLA

---

_{You can retrigger this bot by commenting recheck in this Pull Request. Posted by the CLA Assistant Lite bot.}

[agent-platform-auto-pr]

Go Package Import Differences

Baseline: 6ad2f1d
Comparison: f9ba8b9

binary	os	arch	change
agent	linux	amd64	+1, -0 +github.com/DataDog/datadog-agent/pkg/cli/heuristic
agent	linux	arm64	+1, -0 +github.com/DataDog/datadog-agent/pkg/cli/heuristic
agent	windows	amd64	+1, -0 +github.com/DataDog/datadog-agent/pkg/cli/heuristic
agent	darwin	amd64	+1, -0 +github.com/DataDog/datadog-agent/pkg/cli/heuristic
agent	darwin	arm64	+1, -0 +github.com/DataDog/datadog-agent/pkg/cli/heuristic
iot-agent	linux	amd64	+1, -0 +github.com/DataDog/datadog-agent/pkg/cli/heuristic
iot-agent	linux	arm64	+1, -0 +github.com/DataDog/datadog-agent/pkg/cli/heuristic
heroku-agent	linux	amd64	+1, -0 +github.com/DataDog/datadog-agent/pkg/cli/heuristic
cluster-agent	linux	amd64	+1, -0 +github.com/DataDog/datadog-agent/pkg/cli/heuristic
cluster-agent	linux	arm64	+1, -0 +github.com/DataDog/datadog-agent/pkg/cli/heuristic

[agent-platform-auto-pr]

Files inventory check summary

File checks results against ancestor 6ad2f1d1:

Results for datadog-agent_7.78.0~devel.git.683.f9ba8b9.pipeline.103289181-1_amd64.deb:

No change detected

[agent-platform-auto-pr]

Static quality checks

✅ Please find below the results from static quality gates
Comparison made with ancestor 6ad2f1d
📊 Static Quality Gates Dashboard
🔗 SQG Job

Successful checks

Info

	Quality gate	Change	Size (prev → curr → max)
✅	agent_deb_amd64	+16.12 KiB (0.00% increase)	748.919 → 748.935 → 750.720
✅	agent_deb_amd64_fips	+16.12 KiB (0.00% increase)	707.285 → 707.300 → 711.230
✅	agent_heroku_amd64	+16.12 KiB (0.01% increase)	312.824 → 312.840 → 320.580
✅	agent_msi	+15.0 KiB (0.00% increase)	602.495 → 602.510 → 1062.520
✅	agent_rpm_amd64	+16.12 KiB (0.00% increase)	748.903 → 748.919 → 750.690
✅	agent_rpm_amd64_fips	+16.12 KiB (0.00% increase)	707.268 → 707.284 → 711.210
✅	agent_rpm_arm64	+16.12 KiB (0.00% increase)	727.410 → 727.426 → 732.890
✅	agent_rpm_arm64_fips	+16.12 KiB (0.00% increase)	688.726 → 688.742 → 694.440
✅	agent_suse_amd64	+16.12 KiB (0.00% increase)	748.903 → 748.919 → 750.690
✅	agent_suse_amd64_fips	+16.12 KiB (0.00% increase)	707.268 → 707.284 → 711.210
✅	agent_suse_arm64	+16.12 KiB (0.00% increase)	727.410 → 727.426 → 732.890
✅	agent_suse_arm64_fips	+16.12 KiB (0.00% increase)	688.726 → 688.742 → 694.440
✅	docker_agent_amd64	+16.12 KiB (0.00% increase)	809.250 → 809.266 → 813.040
✅	docker_agent_arm64	+16.12 KiB (0.00% increase)	812.559 → 812.574 → 819.570
✅	docker_agent_jmx_amd64	+16.13 KiB (0.00% increase)	1000.165 → 1000.181 → 1003.920
✅	docker_agent_jmx_arm64	+16.13 KiB (0.00% increase)	992.253 → 992.269 → 999.170
✅	docker_cluster_agent_amd64	+8.06 KiB (0.00% increase)	203.688 → 203.696 → 206.270
✅	iot_agent_deb_amd64	+16.12 KiB (0.04% increase)	43.065 → 43.081 → 43.290
✅	iot_agent_deb_arm64	+16.12 KiB (0.04% increase)	40.124 → 40.140 → 40.920
✅	iot_agent_deb_armhf	+16.06 KiB (0.04% increase)	40.860 → 40.876 → 41.030
✅	iot_agent_rpm_amd64	+16.12 KiB (0.04% increase)	43.066 → 43.082 → 43.290
✅	iot_agent_suse_amd64	+16.12 KiB (0.04% increase)	43.066 → 43.082 → 43.290

9 successful checks with minimal change (< 2 KiB)

	Quality gate	Current Size
✅	docker_cluster_agent_arm64	218.167 MiB
✅	docker_cws_instrumentation_amd64	7.142 MiB
✅	docker_cws_instrumentation_arm64	6.689 MiB
✅	docker_dogstatsd_amd64	39.207 MiB
✅	docker_dogstatsd_arm64	37.445 MiB
✅	dogstatsd_deb_amd64	29.847 MiB
✅	dogstatsd_deb_arm64	28.003 MiB
✅	dogstatsd_rpm_amd64	29.847 MiB
✅	dogstatsd_suse_amd64	29.847 MiB

On-wire sizes (compressed)

	Quality gate	Change	Size (prev → curr → max)
✅	agent_deb_amd64	+54.69 KiB (0.03% increase)	174.183 → 174.237 → 177.700
✅	agent_deb_amd64_fips	-10.33 KiB (0.01% reduction)	165.175 → 165.165 → 172.230
✅	agent_heroku_amd64	+2.86 KiB (0.00% increase)	74.899 → 74.902 → 79.970
✅	agent_msi	+8.0 KiB (0.01% increase)	137.719 → 137.727 → 146.220
✅	agent_rpm_amd64	+8.82 KiB (0.00% increase)	177.116 → 177.125 → 180.780
✅	agent_rpm_amd64_fips	neutral	167.174 MiB → 173.370
✅	agent_rpm_arm64	+25.57 KiB (0.02% increase)	159.632 → 159.657 → 161.610
✅	agent_rpm_arm64_fips	-37.96 KiB (0.02% reduction)	151.281 → 151.244 → 155.910
✅	agent_suse_amd64	+8.82 KiB (0.00% increase)	177.116 → 177.125 → 180.780
✅	agent_suse_amd64_fips	neutral	167.174 MiB → 173.370
✅	agent_suse_arm64	+25.57 KiB (0.02% increase)	159.632 → 159.657 → 161.610
✅	agent_suse_arm64_fips	-37.96 KiB (0.02% reduction)	151.281 → 151.244 → 155.910
✅	docker_agent_amd64	+16.89 KiB (0.01% increase)	267.206 → 267.223 → 271.240
✅	docker_agent_arm64	+9.63 KiB (0.00% increase)	254.475 → 254.485 → 259.800
✅	docker_agent_jmx_amd64	+25.05 KiB (0.01% increase)	335.850 → 335.874 → 339.870
✅	docker_agent_jmx_arm64	-2.0 KiB (0.00% reduction)	319.125 → 319.123 → 324.390
✅	docker_cluster_agent_amd64	neutral	71.289 MiB → 72.920
✅	docker_cluster_agent_arm64	-4.2 KiB (0.01% reduction)	66.926 → 66.921 → 68.220
✅	docker_cws_instrumentation_amd64	neutral	2.999 MiB → 3.330
✅	docker_cws_instrumentation_arm64	neutral	2.729 MiB → 3.090
✅	docker_dogstatsd_amd64	neutral	15.155 MiB → 15.820
✅	docker_dogstatsd_arm64	neutral	14.478 MiB → 14.830
✅	dogstatsd_deb_amd64	neutral	7.886 MiB → 8.790
✅	dogstatsd_deb_arm64	neutral	6.772 MiB → 7.710
✅	dogstatsd_rpm_amd64	neutral	7.896 MiB → 8.800
✅	dogstatsd_suse_amd64	neutral	7.896 MiB → 8.800
✅	iot_agent_deb_amd64	+8.52 KiB (0.07% increase)	11.349 → 11.357 → 12.040
✅	iot_agent_deb_arm64	+4.39 KiB (0.04% increase)	9.664 → 9.668 → 10.450
✅	iot_agent_deb_armhf	+4.76 KiB (0.05% increase)	9.896 → 9.901 → 10.620
✅	iot_agent_rpm_amd64	+2.14 KiB (0.02% increase)	11.370 → 11.372 → 12.060
✅	iot_agent_suse_amd64	+2.14 KiB (0.02% increase)	11.370 → 11.372 → 12.060

[cit-pr-commenter-54b7da]

Regression Detector

Regression Detector Results

Metrics dashboard
Target profiles
Run ID: 3e76d034-d7e7-4394-90bb-db4d0b71cd98

Baseline: 6ad2f1d
Comparison: f9ba8b9
Diff

Optimization Goals: ✅ No significant changes detected

Experiments ignored for regressions

Regressions in experiments with settings containing erratic: true are ignored.

perf	experiment	goal	Δ mean %	Δ mean % CI	trials	links
➖	docker_containers_cpu	% cpu utilization	+2.26	[-0.80, +5.32]	1	Logs

Fine details of change detection per experiment

perf	experiment	goal	Δ mean %	Δ mean % CI	trials	links
➖	docker_containers_cpu	% cpu utilization	+2.26	[-0.80, +5.32]	1	Logs
➖	quality_gate_logs	% cpu utilization	+1.71	[+0.09, +3.34]	1	Logs bounds checks dashboard
➖	tcp_syslog_to_blackhole	ingress throughput	+0.93	[+0.78, +1.08]	1	Logs
➖	ddot_metrics_sum_cumulativetodelta_exporter	memory utilization	+0.60	[+0.38, +0.83]	1	Logs
➖	docker_containers_memory	memory utilization	+0.36	[+0.26, +0.47]	1	Logs
➖	ddot_metrics	memory utilization	+0.34	[+0.16, +0.52]	1	Logs
➖	otlp_ingest_logs	memory utilization	+0.32	[+0.22, +0.41]	1	Logs
➖	quality_gate_metrics_logs	memory utilization	+0.24	[+0.01, +0.48]	1	Logs bounds checks dashboard
➖	tcp_dd_logs_filter_exclude	ingress throughput	+0.00	[-0.10, +0.10]	1	Logs
➖	uds_dogstatsd_to_api_v3	ingress throughput	-0.00	[-0.20, +0.20]	1	Logs
➖	file_to_blackhole_500ms_latency	egress throughput	-0.01	[-0.40, +0.39]	1	Logs
➖	uds_dogstatsd_to_api	ingress throughput	-0.01	[-0.21, +0.19]	1	Logs
➖	file_tree	memory utilization	-0.05	[-0.10, +0.00]	1	Logs
➖	file_to_blackhole_0ms_latency	egress throughput	-0.06	[-0.57, +0.45]	1	Logs
➖	file_to_blackhole_100ms_latency	egress throughput	-0.06	[-0.14, +0.02]	1	Logs
➖	file_to_blackhole_1000ms_latency	egress throughput	-0.10	[-0.53, +0.33]	1	Logs
➖	uds_dogstatsd_20mb_12k_contexts_20_senders	memory utilization	-0.13	[-0.19, -0.07]	1	Logs
➖	ddot_metrics_sum_delta	memory utilization	-0.16	[-0.34, +0.02]	1	Logs
➖	quality_gate_idle	memory utilization	-0.25	[-0.30, -0.20]	1	Logs bounds checks dashboard
➖	ddot_metrics_sum_cumulative	memory utilization	-0.35	[-0.49, -0.21]	1	Logs
➖	ddot_logs	memory utilization	-0.43	[-0.49, -0.37]	1	Logs
➖	otlp_ingest_metrics	memory utilization	-0.45	[-0.60, -0.29]	1	Logs
➖	quality_gate_idle_all_features	memory utilization	-0.56	[-0.59, -0.52]	1	Logs bounds checks dashboard

Bounds Checks: ✅ Passed

perf	experiment	bounds_check_name	replicates_passed	observed_value	links
✅	docker_containers_cpu	simple_check_run	10/10	700 ≥ 26
✅	docker_containers_memory	memory_usage	10/10	275.22MiB ≤ 370MiB
✅	docker_containers_memory	simple_check_run	10/10	693 ≥ 26
✅	file_to_blackhole_0ms_latency	memory_usage	10/10	0.19GiB ≤ 1.20GiB
✅	file_to_blackhole_0ms_latency	missed_bytes	10/10	0B = 0B
✅	file_to_blackhole_1000ms_latency	memory_usage	10/10	0.23GiB ≤ 1.20GiB
✅	file_to_blackhole_1000ms_latency	missed_bytes	10/10	0B = 0B
✅	file_to_blackhole_100ms_latency	memory_usage	10/10	0.20GiB ≤ 1.20GiB
✅	file_to_blackhole_100ms_latency	missed_bytes	10/10	0B = 0B
✅	file_to_blackhole_500ms_latency	memory_usage	10/10	0.22GiB ≤ 1.20GiB
✅	file_to_blackhole_500ms_latency	missed_bytes	10/10	0B = 0B
✅	quality_gate_idle	intake_connections	10/10	3 = 3	bounds checks dashboard
✅	quality_gate_idle	memory_usage	10/10	172.66MiB ≤ 175MiB	bounds checks dashboard
✅	quality_gate_idle_all_features	intake_connections	10/10	2 ≤ 3	bounds checks dashboard
✅	quality_gate_idle_all_features	memory_usage	10/10	489.04MiB ≤ 550MiB	bounds checks dashboard
✅	quality_gate_logs	intake_connections	10/10	3 ≤ 6	bounds checks dashboard
✅	quality_gate_logs	memory_usage	10/10	206.14MiB ≤ 220MiB	bounds checks dashboard
✅	quality_gate_logs	missed_bytes	10/10	0B = 0B	bounds checks dashboard
✅	quality_gate_metrics_logs	cpu_usage	10/10	357.09 ≤ 2000	bounds checks dashboard
✅	quality_gate_metrics_logs	intake_connections	10/10	4 ≤ 6	bounds checks dashboard
✅	quality_gate_metrics_logs	memory_usage	10/10	400.55MiB ≤ 475MiB	bounds checks dashboard
✅	quality_gate_metrics_logs	missed_bytes	10/10	0B = 0B	bounds checks dashboard

Explanation

Confidence level: 90.00%
Effect size tolerance: |Δ mean %| ≥ 5.00%

Performance changes are noted in the perf column of each table:

• ✅ = significantly better comparison variant performance
• ❌ = significantly worse comparison variant performance
• ➖ = no significant change in performance

A regression test is an A/B test of target performance in a repeatable rig, where "performance" is measured as "comparison variant minus baseline variant" for an optimization goal (e.g., ingress throughput). Due to intrinsic variability in measuring that goal, we can only estimate its mean value for each experiment; we report uncertainty in that value as a 90.00% confidence interval denoted "Δ mean % CI".

For each experiment, we decide whether a change in performance is a "regression" -- a change worth investigating further -- if all of the following criteria are true:

1. Its estimated |Δ mean %| ≥ 5.00%, indicating the change is big enough to merit a closer look.

2. Its 90.00% confidence interval "Δ mean % CI" does not contain zero, indicating that if our statistical model is accurate, there is at least a 90.00% chance there is a difference in performance between baseline and comparison variants.

3. Its configuration does not mark it "erratic".

CI Pass/Fail Decision

✅ Passed. All Quality Gates passed.

• quality_gate_idle, bounds check memory_usage: 10/10 replicas passed. Gate passed.
• quality_gate_idle, bounds check intake_connections: 10/10 replicas passed. Gate passed.
• quality_gate_metrics_logs, bounds check missed_bytes: 10/10 replicas passed. Gate passed.
• quality_gate_metrics_logs, bounds check intake_connections: 10/10 replicas passed. Gate passed.
• quality_gate_metrics_logs, bounds check cpu_usage: 10/10 replicas passed. Gate passed.
• quality_gate_metrics_logs, bounds check memory_usage: 10/10 replicas passed. Gate passed.
• quality_gate_logs, bounds check missed_bytes: 10/10 replicas passed. Gate passed.
• quality_gate_logs, bounds check memory_usage: 10/10 replicas passed. Gate passed.
• quality_gate_logs, bounds check intake_connections: 10/10 replicas passed. Gate passed.
• quality_gate_idle_all_features, bounds check memory_usage: 10/10 replicas passed. Gate passed.
• quality_gate_idle_all_features, bounds check intake_connections: 10/10 replicas passed. Gate passed.