metrics-reporter
Skill15 starsChanged 5 months agoMITView on GitHub
Record agentic task completion via event files for DORA observability.
In this repository29 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- llms-full.txtllms.txt
- llms.txtllms.txt
- anti-ai-slopSkill
- architecture-diagramSkill
- atomic-commitSkill
- build-rustSkill
- codacySkill
- crates-io-name-checkSkill
- dora-reportSkill
- goap-agentSkill
- harnessSkill
- issue-triageSkill
- lint-rustSkill
- privacy-firstSkill
- release-rustSkill
- reviewing-pull-requestsSkill
- secret-lintSkill
- self-fix-loopSkill
- skill-creatorSkill
- skill-evaluatorSkill
- stacked-prsSkill
- task-decompositionSkill
- test-rustSkill
- tokio-performanceSkill
- triz-analysisSkill
- triz-solverSkill
- verify-actionsSkill
The file
We split this skill into the rules it gives an agent
6 rules. About 0 are common, 5 are distinctive, and 1 is very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈5
Distinctive · estimate
This author's own choices. Worth a look.
1
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 20
Call this skill as the LAST step of any task or sub-task.
- Distinctive: line 24
Prepare Directory: Ensure the daily directory exists.
- Distinctive: line 31
Generate Filename: Create a unique filename for the event.
- Distinctive: line 53
Verify: Ensure the JSON is valid and written to the correct location.
- Very specific: line 57
When using the stacked-prs skill, add these optional fields to the event JSON:
- Distinctive: line 71
Do NOT include these fields for non-stacked work.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Skill: metrics-reporter
Purpose
Record agentic task completion by writing a per-event JSON file to .agents/events/. This pattern avoids Git merge conflicts and supports multi-agent coordination.
When to use
Call this skill as the LAST step of any task or sub-task.
Steps
Prepare Directory: Ensure the daily directory exists.
EVENT_DIR=".agents/events/$(date -u +%Y/%m/%d)" mkdir -p "${EVENT_DIR}"Generate Filename: Create a unique filename for the event.
EVENT_FILE="${EVENT_DIR}/$(date -u +%Y-%m-%dT%H-%M-%SZ)-${AGENT_NAME:-unknown}-${TASK_ID:-$(date -u +%s)}.json"Write Event Data: Populated with task metadata and success status.
{ "event_id": "$(date -u +%s)-${AGENT_NAME}", "task_id": "${TASK_ID}", "skill": "${SKILL_NAME}", "status": "success", "started_at": "${STARTED_AT}", "finished_at": "$(date -u +%Y-%m-%dT%H:%M:%SZ)", "success": true, "human_interventions": 0, "git_sha": "$(git rev-parse --short HEAD)" }- Verify: Ensure the JSON is valid and written to the correct location.
Stacked PR Metrics
When using the stacked-prs skill, add these optional fields to the event JSON:
{
"stack_id": "billing-feature",
"stack_position": 1,
"stack_total": 3
}stack_id: Identifier for the stack (feature name)stack_position: This layer's position in the stack (1-indexed)stack_total: Total number of layers in the stack
Do NOT include these fields for non-stacked work.
Human Interventions
Set human_interventions > 0 if:
- A human corrected, rewrote, or reverted your code.
- A PR required a fixup commit after your submission.
- A review comment called out an error in your work.
Aggregation
Aggregated metrics are periodically generated by CI into .agents/aggregated/. The CI status files are in .agents/ci/. Do not hand-edit files in these directories.
Rationalizations
| Rationalization | Reality |
| "I'll skip metrics — it's just paperwork." | Without metrics, you can't prove AI tooling ROI or identify bottlenecks. |
| "I'll log it later." | Later never happens. Log immediately after task completion. |
| "human_interventions doesn't matter." | It's the key metric for measuring true agent autonomy vs human effort. |
Red Flags
- Not calling metrics-reporter as the last step of a task
- Setting human_interventions to 0 when a human fixed your code
- Using the old append-to-metrics.jsonl pattern instead of event files
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

