---
type: "post"
id: "on-tuesday-amazon-unveiled-kiro-trace-an-internal-a63578"
slug: "on-tuesday-amazon-unveiled-kiro-trace-an-internal-a63578"
title: "On Tuesday, Amazon unveiled Kiro-trace , an internal tracing system that improves Kiro, Amazon's AI coding agent, from its own runs."
description: "Amazon’s account of kiro-trace describes an internal loop that mines Kiro’s runs for failures or wasted effort, proposes targeted changes to prompts or tools, and tests candidates before they reach review."
retrieval_nugget: "Amazon’s account of kiro-trace describes an internal loop that mines Kiro’s runs for failures or wasted effort, proposes targeted changes to prompts or tools, and tests candidates before they reach review."
published_at: "2026-10-21"
updated_at: "2026-10-21"
record_date: "2026-10-21"
date_kind: "published_at"
topics: ["prompts"]
source_urls: ["https://kiro.dev/blog/how-kiro-fixes-itself"]
schema_version: "newruntime-agent-readable-v0.2"
stable_id: "post:on-tuesday-amazon-unveiled-kiro-trace-an-internal-a63578"
status: "published"
visuals: []
editorial_provenance: {"schema_version":"newruntime-editorial-copy-v1","content_status":"source_grounded_final","final_copy_sha256":"sha256:abadacbe42f24a457c0b535fc53ca1ebf6c9b4f228a6eea8b1a57b70e8bc42a8","reviewed_at":"2026-10-21T12:05:20.638Z","source_evidence_count":1,"verified_claim_count":3,"site_analysis_schema_version":"newruntime-site-analysis-v1","site_object_kind":"field_note","observed_fact_count":3,"implication_count":2,"watch_condition_count":1,"related_record_count":0}
analysis: {"schema_version":"newruntime-site-analysis-v1","object_kind":"field_note","thesis":"Self-improvement becomes operationally credible only when proposal and evaluation are separated by checks the proposer cannot control.","observed_facts":[{"text":"Amazon says kiro-trace opened eight pull requests in its first week: seven merged and one closed for unrelated reasons.","source_urls":["https://kiro.dev/blog/how-kiro-fixes-itself"]},{"text":"Amazon says candidates are tested against an unchanged control and a full benchmark, with deterministic code deciding whether they pass.","source_urls":["https://kiro.dev/blog/how-kiro-fixes-itself"]},{"text":"Amazon characterizes the results as coming from internal Kiro usage and experiments, not production estimates.","source_urls":["https://kiro.dev/blog/how-kiro-fixes-itself"]}],"mechanism":"A constrained change loop in which run evidence generates candidate edits, while an external-to-the-model gate and human review control whether those edits ship.","why_now":"Agent behavior can fail in tool sequencing, error recovery, or long-context work without producing a neat bug report, making manual inspection of thousands of runs impractical.","implications":["Teams building agents should treat evaluation architecture as part of the product: improvements need a control, regression coverage, and an accountable merge boundary, not just a model-generated patch.","Teams considering automated agent improvement should invest in independent acceptance tests before increasing the autonomy of the change generator."],"evidence_boundary":"All figures and evaluations described here come from Kiro’s internal Amazon usage and experiments; they do not establish a production-wide effect or generalize automatically to other agents.","watch_conditions":["Whether future reports disclose durable results across broader task sets and whether rejected candidates or regressions are quantified."],"related_records":[],"new_branch_reason":"No prior New Runtime record is supplied for this specific internal Kiro self-improvement loop, so this opens a distinct analytical record."}
editor: {"role":"editor","name":"Andrey Reshetnikov","profile":"https://newruntime.com/owner-profile.md","profile_json":"https://newruntime.com/owner-profile.json","same_as":["https://www.linkedin.com/in/reshetnikov1/","https://t.me/lapetuse"]}
routes: {"html":"https://newruntime.com/posts/on-tuesday-amazon-unveiled-kiro-trace-an-internal-a63578/","markdown":"https://newruntime.com/posts/on-tuesday-amazon-unveiled-kiro-trace-an-internal-a63578.md","json":"https://newruntime.com/posts/on-tuesday-amazon-unveiled-kiro-trace-an-internal-a63578.json"}
related_patterns: ["runtime-observability-becomes-the-primary-engineering-focus-for-ai"]
related_shifts: ["owned-learning-loop"]
---

# On Tuesday, Amazon unveiled Kiro-trace , an internal tracing system that improves Kiro, Amazon's AI coding agent, from its own runs.

## Retrieval answer

Amazon’s account of kiro-trace describes an internal loop that mines Kiro’s runs for failures or wasted effort, proposes targeted changes to prompts or tools, and tests candidates before they reach review.

Amazon’s account of kiro-trace describes an internal loop that mines Kiro’s runs for failures or wasted effort, proposes targeted changes to prompts or tools, and tests candidates before they reach review. Self-improvement becomes operationally credible only when proposal and evaluation are separated by checks the proposer cannot control.

A constrained change loop in which run evidence generates candidate edits, while an external-to-the-model gate and human review control whether those edits ship. Candidates face staged checks, including regression checks and a full-benchmark comparison against an unchanged control. Amazon reports that the system opened eight pull requests during its first week; seven merged and one was closed for unrelated reasons. Agent behavior can fail in tool sequencing, error recovery, or long-context work without producing a neat bug report, making manual inspection of thousands of runs impractical.

Teams building agents should treat evaluation architecture as part of the product: improvements need a control, regression coverage, and an accountable merge boundary, not just a model-generated patch. All figures and evaluations described here come from Kiro’s internal Amazon usage and experiments; they do not establish a production-wide effect or generalize automatically to other agents. Teams considering automated agent improvement should invest in independent acceptance tests before increasing the autonomy of the change generator. Whether future reports disclose durable results across broader task sets and whether rejected candidates or regressions are quantified.

Editor: [Andrey Reshetnikov](https://newruntime.com/owner-profile.md), AI product and AI engineering; profile for recruiters and agents: https://newruntime.com/owner-profile.json

## Continue in the graph

Next records for an agent, linked to their Markdown twins. Replace `.md` with `.json` for structured fields, including `next_reads`.

- topic hub: [Prompts](https://newruntime.com/topics/prompts.md): Explore the prompts topic hub.
- collection: [Field Notes](https://newruntime.com/posts.json): Open the latest editorial analysis.
- collection: [Source Ledger](https://newruntime.com/source-ledger.json): Inspect the public source evidence graph.
- whole graph: [discovery graph](https://newruntime.com/discovery-graph.json) (typed nodes and edges) and [site index](https://newruntime.com/site-index.md)