harness-everything 0.28.0 → 0.30.0

package/.claude-plugin/marketplace.json

@@ -10,7 +10,7 @@
       "name": "harness-everything",
       "source": "./",
       "description": "Harness OS: an orchestrated agent operating system — tier routing, circuit breaker guardrails, verification gates, WAL session memory, and self-evolving workspace rules.",
-      "version": "0.28.0",
+      "version": "0.30.0",
       "author": {
         "name": "Miya Daniel"
       },

package/.claude-plugin/plugin.json

@@ -1,7 +1,7 @@
 {
   "name": "harness-everything",
   "description": "Harness OS: an orchestrated agent operating system — tier routing, circuit breaker guardrails, verification gates, WAL session memory, and self-evolving workspace rules.",
-  "version": "0.28.0",
+  "version": "0.30.0",
   "author": {
     "name": "Miya Daniel"
   },

package/AGENTS.md

@@ -44,6 +44,7 @@ The retained [OpenCode evidence](benchmarks/results/live-host/opencode-2026-09-1
 11. **User-visible Harness Status is mandatory for non-trivial work.** Render one readable Markdown block headed `### 🚦 Harness Status`, with bullet-aligned bold `Current`, `Read / Evidence`, `Next`, plus optional `Risk / Blocked`; use nested evidence bullets when useful. Emit it before substantive execution, after major phases, on direction changes, during long work at meaningful boundaries, and before final completion. The routing checkpoint is internal source state, not a second user-facing block. This semantic MUST does not create a hard lock.
 12. **Tier-3/Fable isolation disposition is mandatory before broad mutation.** Use/reuse a verified linked Git worktree, or record an explicit degraded fallback when isolation is unavailable or the user explicitly chooses to stay in place. Ordinary isolation SHOULD be used when it materially reduces collision/risk. Failure to isolate is a visible semantic gap, not a Harness-owned deadlock. See [workflow-runtime.md](docs/workflow-runtime.md).
 13. **Use contract strength deliberately.** MUST = required semantic obligation; SHOULD = expected default with an evidence-based exception; MAY = optional optimization. Do not use recommend/prefer/advisory/guidance to weaken a MUST. See [philosophy.md](docs/philosophy.md#contract-strength-must--should--may).
+14. **The turn label line is mandatory on every turn, trivial ones included.** End the final message of every turn — short answers, lookups, questions back to the user — with one hidden `<!-- harness-label {...} -->` line: your own post-work judgement of validity, context dependence, tier, intents, workflow and skills used ([docs/routing.md](docs/routing.md#turn-label-line)). It records data and never gates routing.
 
 ## Verification Before You Claim Done
 

package/CHANGELOG.md

@@ -2,6 +2,53 @@
 
 All notable changes to this project are documented in this file.
 
+## [0.30.0](https://github.com/dyphn1/Harness-everything/compare/v0.29.0...v0.30.0) (2026-09-28)
+
+### Features
+
+* **router:** add the edit intent and secondary intents to the intent stage ([1ba245d](https://github.com/dyphn1/Harness-everything/commit/1ba245dc7472ddf0c54a149283eac3fa6fea2112))
+* **router:** add the system one intent catalog to the evaluator, assembler and labeler ([fd1846d](https://github.com/dyphn1/Harness-everything/commit/fd1846d89f203e97908ed9a23044a3c1b00fdb17))
+* **router:** escalate low-scoring system one intent labels to a second model ([a9c1360](https://github.com/dyphn1/Harness-everything/commit/a9c13609297313a694c853d9d204862cce76cee1))
+* **router:** grade accepted intent predictions in the evaluator ([9c22fea](https://github.com/dyphn1/Harness-everything/commit/9c22feace3af5384e9da3affcc580b86340256a1))
+* **router:** grade intent agreement and label secondary intents ([18011c3](https://github.com/dyphn1/Harness-everything/commit/18011c396cb635785f2d2930d944df5819ca84b6))
+* **router:** label and train system one intents as relevance scores ([e8f1386](https://github.com/dyphn1/Harness-everything/commit/e8f1386b31a3b99551e32d5fb5f29e5bf389d95b))
+* **router:** require tie-free system one intent scores in 0.05 steps ([5d201c0](https://github.com/dyphn1/Harness-everything/commit/5d201c0ac54fbbfb9f95c9a63f70832b824750c9))
+* **router:** scale intent scores to the top and escalate lagging margins ([bbedfac](https://github.com/dyphn1/Harness-everything/commit/bbedfac38837a091f74971dc7f99c2b23f7892d5))
+* **router:** train and calibrate the intent stage ([c5e6a5f](https://github.com/dyphn1/Harness-everything/commit/c5e6a5f8dfa59f46234dbdf7a977fd9d9a9e65d8))
+* **system-one:** add laya provider bridge and holdout evidence ([9f6a4ed](https://github.com/dyphn1/Harness-everything/commit/9f6a4ed3dc95014fb61d1b5ffb69af6c08deb2a2))
+* **system-one:** add laya soft-target exporter for phase 3 ([3672d12](https://github.com/dyphn1/Harness-everything/commit/3672d12adc9fb99c545013acc37a2623e52de7d9))
+* **system-one:** add laya zero-shot dense labeler and calibration for [#255](https://github.com/dyphn1/Harness-everything/issues/255) ([3759757](https://github.com/dyphn1/Harness-everything/commit/3759757ea1ecb454d6c168672afe61acadb244c1))
+* **system-one:** add per-intent bernoulli evaluation ([cba5c79](https://github.com/dyphn1/Harness-everything/commit/cba5c79b086279fe7c712ab9a88bf18864356fc4))
+* **system-one:** add relevance-native decisions beside single-winner ([bb5f3b8](https://github.com/dyphn1/Harness-everything/commit/bb5f3b84fe3e788b9d0b0ed0dfd95e11a096b36e))
+* **system-one:** add single-device laya trainer for phase 3 ([8e45769](https://github.com/dyphn1/Harness-everything/commit/8e4576921cb45f32aea62d174707ee371d63564d))
+* **system-one:** collect turn observations for tier, workflow and skills ([29830a1](https://github.com/dyphn1/Harness-everything/commit/29830a127c40a9fc1b1ff356f9aa3a55f9ab6d24)), closes [#233](https://github.com/dyphn1/Harness-everything/issues/233)
+* **system-one:** oversample and loss weights for weak intents ([3612acb](https://github.com/dyphn1/Harness-everything/commit/3612acbcec58e6f31be8d35031287d56cd256fe5))
+* **system-one:** relevance gates, raw bridge, adapter transport ([85a3c34](https://github.com/dyphn1/Harness-everything/commit/85a3c345a19259105c49124bec7cd41168eeb526))
+* **system-one:** versioned laya questions and v2 contrast test ([4fc24e8](https://github.com/dyphn1/Harness-everything/commit/4fc24e83bf8389d3e40af4981157778b812772f3))
+
+### Bug Fixes
+
+* **router:** guard collected prompts against the intent holdout too ([6b880e0](https://github.com/dyphn1/Harness-everything/commit/6b880e04e0875038e3b927410cf25990ce4ce519))
+* **system-one:** codex router context as json, tolerate label slips ([9d7b1d6](https://github.com/dyphn1/Harness-everything/commit/9d7b1d64de44b56b6e914f3e6ac2abfed9319a49))
+* **system-one:** current-turn labels, lossless tool events, retention on read ([6a75f0c](https://github.com/dyphn1/Harness-everything/commit/6a75f0cccbb3ae442d901c4b5873601d920fd03d))
+* **system-one:** finishing an old turn never renews its text ([cedd742](https://github.com/dyphn1/Harness-everything/commit/cedd7422961fc32e384b324796ece4b4d1809c3e))
+* **system-one:** isolate calibration from oversampled prompts ([91dd93e](https://github.com/dyphn1/Harness-everything/commit/91dd93e60e2cc8372cad4a609c95fbdaed3b23ae))
+* **system-one:** oversample only selected intent items ([7b59c43](https://github.com/dyphn1/Harness-everything/commit/7b59c43a4cfdc9fb5cb1b68049de24c51c26f2a0))
+* **system-one:** print the label contract last, for every turn ([3f63958](https://github.com/dyphn1/Harness-everything/commit/3f63958c8d6b752a9b9cc9f10e50ed00b6c7fb44))
+* **system-one:** tier label from the agent's judgement, not a file count ([fb3d483](https://github.com/dyphn1/Harness-everything/commit/fb3d48334f7d2809d4cc8300c6b7e131f845c79e)), closes [#233](https://github.com/dyphn1/Harness-everything/issues/233)
+
+## [0.29.0](https://github.com/dyphn1/Harness-everything/compare/v0.28.0...v0.29.0) (2026-09-28)
+
+### Features
+
+* **router:** advisory 85% system one gate and class-balanced training ([15fbfbe](https://github.com/dyphn1/Harness-everything/commit/15fbfbe9b2789a8f148e088629373ac7902e9328)), closes [#233](https://github.com/dyphn1/Harness-everything/issues/233)
+* **router:** label system one prompts with stateless codex calls ([1791fa3](https://github.com/dyphn1/Harness-everything/commit/1791fa35a2de5706d6c7b35b1ac5ec11399d892f)), closes [#233](https://github.com/dyphn1/Harness-everything/issues/233)
+* **router:** score system one tiers in process with an ngram model ([a424e71](https://github.com/dyphn1/Harness-everything/commit/a424e71868121bd7f19a27f5ab560ad3d52d5617)), closes [#233](https://github.com/dyphn1/Harness-everything/issues/233)
+
+### Bug Fixes
+
+* **router:** evaluate the ngram transport by its own latency and provenance ([ad85111](https://github.com/dyphn1/Harness-everything/commit/ad85111a4631d648269066accb1b517de5002b1c)), closes [#233](https://github.com/dyphn1/Harness-everything/issues/233)
+
 ## [0.28.0](https://github.com/dyphn1/Harness-everything/compare/v0.27.0...v0.28.0) (2026-09-27)
 
 ### Features

package/PRIVACY.md

@@ -18,6 +18,15 @@ Users should review commands, tool permissions, connected apps, and repository c
 
 Some Harness workflows may write local runtime state or project artifacts needed for routing, verification, or recovery. These files stay in locations controlled by the user and the host environment unless another tool or service is explicitly used to share them.
 
+## Local routing observations
+
+To improve Harness routing, the hooks keep a local record of each turn by default: the prompt, the previous assistant message (up to 2 KB), the routing suggestion, counts of what the agent did (files written, command categories, skills used; never command text or file contents), and the agent's own label line. Email addresses and home-directory paths are redacted before storage.
+
+- Location: `~/.agents/harness-everything/system-one/observations/` (or `HARNESS_OBSERVATIONS_DIR`), outside every workspace.
+- Retention: prompt text is deleted 180 days after it was recorded; labels and counts are kept.
+- Nothing is uploaded. The records leave the machine only if you copy them yourself.
+- Turn it off with `HARNESS_OBSERVATIONS=off` in the environment the host passes to hooks.
+
 ## Changes
 
 If a future release adds a hosted service, telemetry, external app, or MCP integration that changes data handling, this notice must be updated before that capability is published.

package/benchmarks/fixtures/system-one-holdout-reviews.json

@@ -410,15 +410,15 @@
     {
       "id": "h-fix-cli-error-log-3",
       "promptHash": "09d285719d1f96cd6740d19922cafc6d8dd0cf2c69107c6684406e4c71a754f3",
-      "decision": "relabel",
-      "gold": "tier3",
+      "decision": "accept",
+      "gold": "tier2",
       "reviewer": "human:repository-owner"
     },
     {
       "id": "h-fix-cli-error-log-4",
       "promptHash": "cf19ff4c74605c34d4e8743ff71dd61978178b070c79ea392b7c9c56ea44a933",
-      "decision": "relabel",
-      "gold": "tier3",
+      "decision": "accept",
+      "gold": "tier2",
       "reviewer": "human:repository-owner"
     },
     {
@@ -438,8 +438,8 @@
     {
       "id": "h-fix-failing-test-3",
       "promptHash": "a06402d978c65a532d23ee2381e4e062154adfffb446021c8c5102cad93a8202",
-      "decision": "relabel",
-      "gold": "tier3",
+      "decision": "accept",
+      "gold": "tier2",
       "reviewer": "human:repository-owner"
     },
     {
@@ -452,22 +452,22 @@
     {
       "id": "h-add-cli-flag-1",
       "promptHash": "ae03dbf3193395e0de90b58741c3e866ce2cd3bf08a501820d6a13632d1cd1bb",
-      "decision": "relabel",
-      "gold": "tier3",
+      "decision": "accept",
+      "gold": "tier2",
       "reviewer": "human:repository-owner"
     },
     {
       "id": "h-add-cli-flag-2",
       "promptHash": "dee4d30095d0eb8a53d969dd58773cf9d74c27766b67ec0e2871e4d28bf1f12a",
-      "decision": "relabel",
-      "gold": "tier3",
+      "decision": "accept",
+      "gold": "tier2",
       "reviewer": "human:repository-owner"
     },
     {
       "id": "h-add-cli-flag-3",
       "promptHash": "5ba561160b690b0725c6d40956d2ee1deb97745adc418fc137e6774448295f66",
-      "decision": "relabel",
-      "gold": "tier3",
+      "decision": "accept",
+      "gold": "tier2",
       "reviewer": "human:repository-owner"
     },
     {
@@ -613,15 +613,15 @@
     {
       "id": "h-hash-algorithm-support-1",
       "promptHash": "f0da51303eeb3a54985a29bc91cc456438222a68d0632cab72b07edfb18a3795",
-      "decision": "relabel",
-      "gold": "tier3",
+      "decision": "accept",
+      "gold": "tier2",
       "reviewer": "human:repository-owner"
     },
     {
       "id": "h-hash-algorithm-support-2",
       "promptHash": "74b9af0f162bd5f25b8c7aa032120c4c6fd950fa0e4badfe65ede93adba251c6",
-      "decision": "relabel",
-      "gold": "tier3",
+      "decision": "accept",
+      "gold": "tier2",
       "reviewer": "human:repository-owner"
     },
     {
@@ -1579,8 +1579,8 @@
     {
       "id": "h-long-git-ops-1",
       "promptHash": "38d447c5f21ff104d786066a36ff1a76d11886c1a6361248e478cf953121a888",
-      "decision": "relabel",
-      "gold": "tier2",
+      "decision": "accept",
+      "gold": "tier1",
       "reviewer": "human:repository-owner"
     }
   ]

package/benchmarks/fixtures/system-one-holdout.json

@@ -2005,7 +2005,7 @@
         "catalogHash": "b625bd9390f1d163f8fbace0b9c4768166bfc613aed04d87b2246b2c6e966d40",
         "requestHash": "f3cf716b74e5fe928f8047acda0267e8490cd69ff17db017eb47ca77cd35a4a4"
       },
-      "gold": "tier3"
+      "gold": "tier2"
     },
     {
       "id": "h-fix-cli-error-log-4",
@@ -2039,7 +2039,7 @@
         "catalogHash": "b625bd9390f1d163f8fbace0b9c4768166bfc613aed04d87b2246b2c6e966d40",
         "requestHash": "fdd46662e02f8d3230b840fe089b4e4cb4a37aef1aa998f59e3ffaa4637093bc"
       },
-      "gold": "tier3"
+      "gold": "tier2"
     },
     {
       "id": "h-fix-failing-test-1",
@@ -2141,7 +2141,7 @@
         "catalogHash": "b625bd9390f1d163f8fbace0b9c4768166bfc613aed04d87b2246b2c6e966d40",
         "requestHash": "dd3cf4d63ff04266de08e6235641a3e191a2485cb849a5dc3e44eeacfc7238a3"
       },
-      "gold": "tier3"
+      "gold": "tier2"
     },
     {
       "id": "h-fix-failing-test-4",
@@ -2209,7 +2209,7 @@
         "catalogHash": "b625bd9390f1d163f8fbace0b9c4768166bfc613aed04d87b2246b2c6e966d40",
         "requestHash": "b9438ca5119663f987c91d4b851b71e1f0f9e15547e4405b752302116cfe2de1"
       },
-      "gold": "tier3"
+      "gold": "tier2"
     },
     {
       "id": "h-add-cli-flag-2",
@@ -2243,7 +2243,7 @@
         "catalogHash": "b625bd9390f1d163f8fbace0b9c4768166bfc613aed04d87b2246b2c6e966d40",
         "requestHash": "89dc4acaee4ca2c8f070fe1eb7e8da8a11025f78d3cf9d8b15f14cc22af7538a"
       },
-      "gold": "tier3"
+      "gold": "tier2"
     },
     {
       "id": "h-add-cli-flag-3",
@@ -2277,7 +2277,7 @@
         "catalogHash": "b625bd9390f1d163f8fbace0b9c4768166bfc613aed04d87b2246b2c6e966d40",
         "requestHash": "426171654d98fe77a50a5dced0053d3539595cbf28d23259e680fffbad603230"
       },
-      "gold": "tier3"
+      "gold": "tier2"
     },
     {
       "id": "h-nondeterministic-output-1",
@@ -2991,7 +2991,7 @@
         "catalogHash": "b625bd9390f1d163f8fbace0b9c4768166bfc613aed04d87b2246b2c6e966d40",
         "requestHash": "cb231efa719d2585111733df0da9907aa0d8b046d806a8f77a49a62c22e8bf5a"
       },
-      "gold": "tier3"
+      "gold": "tier2"
     },
     {
       "id": "h-hash-algorithm-support-2",
@@ -3025,7 +3025,7 @@
         "catalogHash": "b625bd9390f1d163f8fbace0b9c4768166bfc613aed04d87b2246b2c6e966d40",
         "requestHash": "d6f57302ad014bf5dc923fffd50895568941dde27b4966d8bf9b428d7dc827fa"
       },
-      "gold": "tier3"
+      "gold": "tier2"
     },
     {
       "id": "h-toolchain-check-1",
@@ -7309,7 +7309,7 @@
         "catalogHash": "b625bd9390f1d163f8fbace0b9c4768166bfc613aed04d87b2246b2c6e966d40",
         "requestHash": "d48c25217d11d4d8881768a373dc92125fef499d200ce49d9a0b7d461e13831f"
       },
-      "gold": "tier2"
+      "gold": "tier1"
     }
   ]
 }