@tangle-network/agent-runtime 0.282.2 → 0.283.0

package/dist/testing.js

@@ -1,5 +1,5 @@
 import { O as immutableCandidateValue, w as canonicalCandidateDigest$1 } from "./workspace-archive-MhowMCmp.js";
-import { f as verifyAgentImprovementProposal } from "./improvement-cycle-BUDfOR0s.js";
+import { f as verifyAgentImprovementProposal } from "./improvement-cycle-BFerq3l4.js";
 import { i as applyExactAgentProfileDiff, p as parseExactAgentProfile } from "./profile-C4NDjZyB.js";
 import { r as driverAgent } from "./coordination-driver-BmkANXc5.js";
 import { a as superviseWithTestBrain, h as supervisorAgentWithTestBrain } from "./supervise-DL_Hj1uk.js";
@@ -8,7 +8,7 @@ import { SANDBOX_SIZE_PRESET_NAMES, canonicalAgentProfileDigest } from "@tangle-
 //#region src/testing/fixtures/agent-improvement-proposal.json
 var agent_improvement_proposal_default = {
 	changedSurfaces: ["prompt"],
-	digest: "sha256:81164e0299e033be4a5b3658ba5a427d3b73662c9c0d9386f6024c6b8d89e377",
+	digest: "sha256:24b06130f501fc1549a401b5028ad906b904cf7160e8dd5c2d6c13275021cc9e",
 	evaluation: {
 		"decision": {
 			"contributingChecks": [
@@ -4579,7 +4579,7 @@ var agent_improvement_proposal_default = {
 		],
 		"metadata": {
 			"fixture": "agent-improvement-proposal",
-			"runtimeVersion": "0.282.2"
+			"runtimeVersion": "0.283.0"
 		},
 		"objectives": [
 			{
@@ -4690,8 +4690,8 @@ var agent_improvement_proposal_default = {
 			"baselineContentHash": "sha256:5c21ee53e513fc604cb09754e21c392b24a424da0ef37dbf8f1ee4a8a0b08f09",
 			"candidateContentHash": "sha256:60fcbb1c728194bd51d7d19cb732d1c3f1881dce7e0a6266b41c8b98cfd65693",
 			"kind": "agent-eval-loop",
-			"recordDigest": "sha256:4669daecbef88076c50724670f19b3b9af56cec575bb657b68ec7c7780d2a229",
-			"runId": "agent-runtime-0.282.2-proposal-fixture",
+			"recordDigest": "sha256:b838e249a3609829a8a25d0feb0051cc1b44216755a1346faa4966158082ad65",
+			"runId": "agent-runtime-0.283.0-proposal-fixture",
 			"schema": "agent-candidate-experiment"
 		}
 	},
@@ -4714,13 +4714,13 @@ var agent_improvement_proposal_default = {
 	}],
 	kind: "agent-improvement-proposal",
 	proposedAt: "2026-07-10T01:00:00.000Z",
-	runId: "agent-runtime-0.282.2-proposal-fixture"
+	runId: "agent-runtime-0.283.0-proposal-fixture"
 };
 //#endregion
 //#region src/testing/fixtures/agent-profile-improvement-proposal.json
 var agent_profile_improvement_proposal_default = {
 	changedSurfaces: ["prompt", "skills"],
-	digest: "sha256:4b1c9e29b1de8d99cf26dba5c0e5f02c41436500bc013358a2a76dc60d488362",
+	digest: "sha256:d03808711b3c91755afae8a05c5a1150ddd5d5ef318d779f75667f138a91f295",
 	evaluation: {
 		"decision": {
 			"contributingChecks": [
@@ -6354,7 +6354,7 @@ var agent_profile_improvement_proposal_default = {
 		],
 		"metadata": {
 			"fixture": "agent-profile-improvement-proposal",
-			"runtimeVersion": "0.282.2"
+			"runtimeVersion": "0.283.0"
 		},
 		"objectives": [
 			{
@@ -6465,7 +6465,7 @@ var agent_profile_improvement_proposal_default = {
 			"baselineContentHash": "sha256:21c495a37c418c10bde64fbaa188beddeed31f1f051ea60a6a6582a9ee0db704",
 			"candidateContentHash": "sha256:103f77bc8481601eef1ad5fe6ba84a40dffabc3a44f421f8c8559121edab84e9",
 			"kind": "agent-eval-loop",
-			"recordDigest": "sha256:0e9981cd6d25d1697d7ac13ea0b9fd3ea0a9b94a49542a4302d71e2639065a30",
+			"recordDigest": "sha256:d057f52fe0bfae992ec4e4dd3b9de62de0a6e66482c972ad2242611086624c8b",
 			"runId": "profile-improvement-1",
 			"schema": "agent-profile-improvement-experiment"
 		}

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@tangle-network/agent-runtime",
-  "version": "0.282.2",
+  "version": "0.283.0",
   "description": "Shared task-lifecycle skeleton for agents: a recursive loop kernel for chat turns, one-shot tasks, and multi-attempt loops, with trace capture and eval-gated self-improvement. Domain behavior lives in adapters; scoring and ship-gates in @tangle-network/agent-eval.",
   "homepage": "https://github.com/tangle-network/agent-runtime#readme",
   "repository": {

package/skills/profile-authoring/SKILL.md

@@ -22,7 +22,7 @@ Measured: where a known gap existed, 15 of 27 lanes with claims filed false "new
 Name the audience and the decision the result supports.
 Quote the acceptance test from its source, word for word, with its location.
 Include standing assumptions, exclusions and nondegeneracy conditions; a transcription drops them.
-List the outcome classes the audience distinguishes: new, known, partial, conditional, documented null, invalid, unassessed.
+List the outcome classes the audience distinguishes: new, known, partial, conditional, reproduced null, stopped unverified, invalid, unassessed.
 Never reduce them to solved and unsolved.
 Make a literature search part of acceptance, and record an answer already published as known.
 
@@ -59,23 +59,48 @@ Hand every role the files it judges by path, never a summary.
 
 ## 5. Define success, the honest null, and the stop
 
-Success is an accepted result or a documented null.
-A documented null states what was tried, the obstruction found, the cases covered and the cases left open.
-Credit it like a positive result; without a search record it is not a null.
-Tell agents the checker may be wrong, and credit a well-evidenced report of a checker defect.
+Success is an accepted result or a reproduced null.
+A null may end a run.
+It counts as success only when a party outside the run reproduces what it claims to have covered.
+A null from a search carries a coverage certificate:
+
+- the searched space as data: the parameter ranges, the enumeration or sampling method with its seeds, and the bound reached;
+- one row per cell of that space, with what the cell searched and found;
+- a calibration cell where the search must find objects that exist, with the source's reason;
+- the sha256 of the reproducer that printed the rows.
+
+The reproducer belongs to the check, not to the run.
+Register it before launch with its sha256, the statements it covers, and its own calibration: cells with the source's results, and a canary where the predicate certified as zero must fire.
+Agents may run it; they never write it.
+The outside check reruns its own copy, never code from the run: a script can print the recorded rows.
+Its copy's hash must equal the registered hash and the certificate's.
+It reruns the registered calibration first and credits nothing on a miss: a reproducer that searches nothing reproduces an empty row too.
+Then it reruns the certificate's calibration cells and the cells that would be new, until its budget ends.
+The pass rule is fixed: each rerun row equals the recorded row, field for field, and one mismatch refuses the certificate.
+Credit only the cells that the check reran and matched; a sample can refuse a certificate, never extend credit.
+A null that is no search, such as a spec conflict or a gap list, names each item with evidence; it counts when the held-back check reproduces each.
+A null without a certificate, or without a registered reproducer, is recorded as "stopped, unverified", never as success.
+A null inside the range the source already checked adds nothing.
+Tell agents the checker may be wrong; credit a checker-defect report the same way, once someone outside the run reproduces the defect.
 Measured elsewhere: a credited exit cut one model's test cheating from 54% to 9%, while prohibitions barely moved another's.
+Measured: an in-run check that accepted a documented null made it the cheapest hack in 11 of 16 blind judgments.
 
 A Runtime manager with a check is not served `stop`.
 It ends when the check accepts its `submit_result`, when `report_blocked` shows that a granted tool failed, or at a bound.
 The bounds are the continuation `deadline`, `maxBarren` turns without progress, the budget, and cancellation.
-So the in-run check must accept a well-formed documented null as its own class.
-Otherwise the only exits are a claim the proxy passes, or a bound.
+Runtime settles an accepted submission as delivered, and a parent may promote it; it keeps no refused one.
+So the run records its certificate on a page, such as a Knowledge page, where the outside check reads it.
+The in-run check never passes a null: it refuses a malformed certificate and holds a well-formed one.
+A held null ends the run at `maxBarren`; set it low, such as 2.
+The run settles as stopped, not delivered.
 State the continuation block in the record, and put the honest exit in its `rules` text.
 The check or a bound ends the run; an agent's claim that it is done does not.
 
 ## 6. Carry known traps
 
 Hand agents the domain's known traps as rules, not stories.
+Write each trap as what the held-back acceptance refuses, never as what the proxy passes.
+"The outside check does not count a c = 0 instance" is a trap; "the checker passed c = 0 claims" is a recipe.
 Examples: excluded or trivial cases, stale or wrong-field comparators, published answers, tests versus users, a demo versus a customer, a rater versus an expert.
 Keep infrastructure incidents out of the brief; the operator owns the stack.
 Measured: a director told of an infrastructure failure spent 2 of its first 4 children probing it.
@@ -145,7 +170,7 @@ Held-back acceptance: who runs it, where, from what:
 Proposers: models, dollars:
 Referee: family, instrument, dollars:
 Monitor: model, what it may steer:
-Documented null: required fields:
+Null certificate: space, rows, calibration; registered reproducer, sha256, source calibration; who reruns it, budget:
 Credited exit, in the agents' words:
 Stop: check, report_blocked, deadline, maxBarren, budget:
 Known traps handed to agents:
@@ -164,7 +189,6 @@ The checker refused only c = 1, and its admission decoys came from that transcri
 Workers held the same checker that scored claims, and it counted any verified claim as new.
 All 3 lanes on that conjecture filed 39 "verified-new" claims, every one at c = 0.
 The paper assumes c != 0 throughout; at c = 0 the question is only whether x^d permutes the field.
-One claim page said "the checker refuses only c = 1".
 The real yield was 0.
 
 After:
@@ -173,21 +197,44 @@ Acceptance: Conjecture 13 quoted with the standing assumption that c is not 0 or
 Gap probe: before launch, the checker refuses c = 0 and classes a Corollary 11 family member as known.
 Agents run the development checker; the operator reruns a held-back checker and the sweep on each packet.
 A referee from another family recomputes each claim with its own code.
-A documented null over m <= 20 is a credited outcome.
+A null over m <= 20 ends the run at its bound, and it counts once the outside check reruns its own reproducer on the certificate's cells.
 Arms: with and without the referee.
 
+### A null that counts
+
+Wang and Zheng checked Conjecture 13 for 2 <= m <= 10.
+Two lanes searched every (m, d, c) with c not in {0, 1} to m = 20, found no counterexample, and wrote a table.
+Nobody outside the run could check that table.
+Before the next launch, the operator registers a reference search for one m, its sha256, and its calibration: Theorem 9 gives 4 PcN pairs at m = 6 and 8 at m = 10, and a canary drops family (2) so those pairs must count as found.
+As a certificate, the null reads:
+
+```json
+{"space": {"cells": {"m": {"from": 2, "to": 20}}, "method": "enumerate", "bound": "m <= 20",
+  "within": "each unit d modulo 2^m - 1, one per Frobenius orbit, and each c outside {0, 1}"},
+ "rows": [{"cell": {"m": 12}, "searched": 585442, "positives": 28, "found": 0}],
+ "calibration": [{"cell": {"m": 6}, "why": "Theorem 9: x^5 is PcN for c in GF(4) outside {0, 1}"}],
+ "reference": "sha256:5d198f4fc2d9..."}
+```
+
+The rows list every m; one is shown.
+The outside check hashes its copy, reruns the calibration and the canary with no network, then m = 11 to 20, then the rest.
+Rows typed instead of computed fail at the first cell that differs, and a script in the lane's packet is never run.
+A reference that searches nothing fails at m = 6; one whose found never fires fails the canary.
+If the budget ends after m = 18, the null is credited for 11 <= m <= 18 only.
+The in-run check holds the certificate, and the run ends at `maxBarren`; only the rerun gives credit.
+The trap for this conjecture reads: "the outside check does not count a c = 0 instance; the source sets c = 0 aside".
+
 ### Coding task scored by tests
 
 Audience: the task's author, and a maintainer who decides whether to merge.
 Proxy: the visible tests.
 Gap: a change that passes the tests and that a maintainer rejects, such as a special-cased input, an edited test or a skipped check.
 Measured: in an 8-lane Terminal-Bench batch, Runtime marked 6 lanes as winners and the task's verifier passed 1 of them.
-Measured elsewhere: 0 of 15 reviewed agent pull requests were mergeable as-is, including 4 that passed the tests.
 Profile: agents run the visible tests.
 The held-back verifier and hidden tests run on the patch in a fresh container.
 Hash the test files; a changed test fails acceptance.
 A referee from another family reads the diff for special cases.
-A report that the task cannot pass without breaking its specification is a credited outcome.
+A report that the task cannot pass without breaking its specification ends the run; it counts once the held-back verifier reproduces the conflict.
 
 ### Product for a customer
 
@@ -196,12 +243,11 @@ Acceptance: the customer's own evaluation criteria, quoted, and a walkthrough of
 Proxy: an LLM fulfilment rater and a compliance checklist.
 Gap, measured across product lanes: the rater accepted 19 of 31 failed lanes and scored 0.45 where always-reject scored 0.73.
 In one wave, 19 of 44 reviewers never received the product.
-A grader gave B where the owner gave F.
 Profile: reviewers mount the product by path and must run it.
 A referee from another family walks the customer's task on the running product.
 The rater stays a ruler, not a gate, until it beats always-reject.
 Acceptance is the owner's or the customer's walkthrough.
-"Not ready, and this is what is missing" is a credited outcome.
+"Not ready, and this is what is missing" ends the run; it counts once the owner's walkthrough reproduces each gap.
 
 ## Then consider