fix: round-5 trial findings and first baseline; release 0.14.10

CHANGELOG.md

@@ -1,5 +1,26 @@
 # Changelog
 
+## [0.14.10] — 2026-09-25
+
+From round 5, the first measured trial round: seven trials on 0.14.9, all isolated. The
+full-package trial generated every ESS output and ran `go test` against the synthesized suite: 13
+passed, 0 failed, 0 skipped.
+
+- `agentplugins-check trial-report` recognises every spelling of `go test` (`go -C <dir> test`,
+  behind environment assignments, after `cd … &&`) and only in command position; it had reported the
+  full-package run as "not run".
+- `b10x setup plan` with no selection keeps every installed product, optional ones included; it
+  proposed uninstalling an installed `connectors`.
+- `ess:specifying`: every `--kind` of `ess generate` (`schema`, `openapi`, `asyncapi`, `docs`,
+  `site`, `docs-ir`) and `ess generate types`; one `--out` per kind side by side, because nested
+  output directories are refused; which kinds add their own folder (`site` does not). OpenAPI and
+  AsyncAPI need a component that says how the service is reached, or they write `0 artifact(s)`
+  without a refusal (beyond10x/ess#102).
+- `ess:testing-conformance`: generate the Go suite inside the implementation's module; return payload
+  numbers as `float64` (beyond10x/ess#101).
+- Trials: `cargo` is allowed (the Rust type library can be compiled); the aep, worktree and upgrade
+  prompts say nobody will answer questions.
+
 ## [0.14.9] — 2026-09-25
 
 Trials become measurable: their prompts live in the repository, and each run is scored against a

Cargo.toml

@@ -3,7 +3,7 @@ resolver = "2"
 members = ["crates/agentplugins-check", "crates/b10x"]
 
 [workspace.package]
-version = "0.14.9"
+version = "0.14.10"
 edition = "2021"
 rust-version = "1.85"
 license = "Apache-2.0"

Taskfile.yml

@@ -91,7 +91,7 @@ tasks:
       - |
         (. '{{.T}}/env' && dir="$(cat '{{.T}}/workdir')" && mkdir -p "{{.T}}/work/$dir" && cd "{{.T}}/work/$dir" &&
          timeout 2400 claude -p "$(cat '{{.T}}/prompt.txt')" --permission-mode acceptEdits \
-           --allowedTools 'Bash(b10x:*)' 'Bash(ess:*)' 'Bash(aep:*)' 'Bash(worktree:*)' 'Bash(git:*)' 'Bash(go:*)' \
+           --allowedTools 'Bash(b10x:*)' 'Bash(ess:*)' 'Bash(aep:*)' 'Bash(worktree:*)' 'Bash(git:*)' 'Bash(go:*)' 'Bash(cargo:*)' \
              'Bash(mkdir:*)' 'Bash(ls:*)' 'Bash(cat:*)' 'Bash(find:*)' 'Bash(curl:*)' 'Bash(tar:*)' \
              'Bash(sha256sum:*)' 'Bash(mv:*)' 'Bash(chmod:*)' Read Write Edit Glob Grep Skill Task Agent \
            --strict-mcp-config --output-format stream-json --verbose > '{{.T}}/run.jsonl' 2> '{{.T}}/run.err'

crates/agentplugins-check/src/report.rs

@@ -283,10 +283,44 @@ fn go_counts(output: &str) -> Option<GoTest> {
     Some(counts)
 }
 
+/// Whether a shell command runs `go test`, however it is spelled: `go test`, `go -C <dir> test`,
+/// behind environment assignments or after `cd … &&`.
+fn runs_go_test(command: &str) -> bool {
+    let words: Vec<String> = shlex::split(command)
+        .unwrap_or_else(|| command.split_whitespace().map(str::to_owned).collect());
+    words.iter().enumerate().any(|(index, word)| {
+        if word != "go" && !word.ends_with("/go") {
+            return false;
+        }
+        // In command position: every word since the last separator is an environment assignment.
+        let start = words[..index]
+            .iter()
+            .rposition(|w| matches!(w.as_str(), "&&" | "||" | ";" | "|" | "then" | "do"))
+            .map_or(0, |separator| separator + 1);
+        if !words[start..index]
+            .iter()
+            .all(|w| w.contains('=') && !w.starts_with('-'))
+        {
+            return false;
+        }
+        let mut rest = words[index + 1..].iter();
+        while let Some(next) = rest.next() {
+            match next.as_str() {
+                "-C" => {
+                    rest.next();
+                }
+                flag if flag.starts_with('-') => {}
+                verb => return verb == "test",
+            }
+        }
+        false
+    })
+}
+
 fn go_test(run: &Run) -> Option<GoTest> {
     run.shells
         .iter()
-        .filter(|shell| shell.command.contains("go test"))
+        .filter(|shell| runs_go_test(&shell.command))
         .rev()
         .find_map(|shell| go_counts(shell.output.as_deref()?))
 }
@@ -657,6 +691,27 @@ mod tests {
         std::fs::remove_dir_all(&sandbox).unwrap();
     }
 
+    #[test]
+    fn every_spelling_of_go_test_is_recognised() {
+        for command in [
+            "go test ./...",
+            "cd impl && go test -v ./...",
+            "go -C /work/impl test -count=1 -v ./... 2>&1 | tail -20",
+            "ESS_REPORT_OUT=r.json go -C impl test ./...",
+            "/usr/local/go/bin/go test ./...",
+        ] {
+            assert!(runs_go_test(command), "{command}");
+        }
+        for command in [
+            "go vet ./...",
+            "go -C impl build ./...",
+            "echo go test",
+            "cargo test",
+        ] {
+            assert!(!runs_go_test(command), "{command}");
+        }
+    }
+
     #[test]
     fn go_test_counts_leaf_tests_of_the_last_run() {
         let sandbox = scratch("gotest");

crates/b10x/src/plan.rs

@@ -270,15 +270,9 @@ pub struct Context<'a> {
 pub fn plan(context: &Context<'_>, inventory: &Inventory) -> Plan {
     let catalog = context.catalog;
     let present = present(catalog, inventory);
-    let selected: BTreeSet<String> = context.selection.clone().unwrap_or_else(|| {
-        present
-            .iter()
-            .filter(|id| {
-                context.only || catalog.product(id).is_some_and(|product| !product.optional)
-            })
-            .cloned()
-            .collect()
-    });
+    // With no explicit selection, what is installed stays installed, optional products included:
+    // leaving one out would uninstall it (connectors, trial 5).
+    let selected: BTreeSet<String> = context.selection.clone().unwrap_or_else(|| present.clone());
     let offers = catalog
         .products
         .iter()
@@ -1360,6 +1354,32 @@ mod tests {
         plan(&context, inventory)
     }
 
+    #[test]
+    fn an_installed_optional_product_stays_when_nothing_is_selected() {
+        let mut claude = HostState::default();
+        claude.marketplaces.push(market());
+        claude.plugins.push(installed("b10x", "b10x", "user", None));
+        claude
+            .plugins
+            .push(installed("connectors", "b10x", "user", None));
+        let inventory = Inventory {
+            claude: Some(claude),
+            ..Inventory::default()
+        };
+        let plan = run(&inventory, None, &[Host::Claude]);
+        let commands = commands(&plan);
+        assert!(
+            !commands
+                .iter()
+                .any(|c| c.contains("uninstall connectors@b10x")),
+            "{commands:#?}"
+        );
+        assert!(plan
+            .offers
+            .iter()
+            .any(|offer| offer.id == "connectors" && offer.selected));
+    }
+
     #[test]
     fn upgrade_installs_the_replacement_of_a_legacy_plugin_it_removes() {
         let mut claude = HostState::default();

plugins/aep/.claude-plugin/plugin.json

@@ -2,7 +2,7 @@
   "name": "aep",
   "displayName": "AEP",
   "description": "Plan governed work in the AEP artifact store and deliver it in reviewed waves: decomposition, plan critique, reverse engineering, story scoping, implementation and adversarial review.",
-  "version": "0.14.9",
+  "version": "0.14.10",
   "author": {
     "name": "Beyond10x"
   },

plugins/aep/.codex-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "aep",
-  "version": "0.14.9",
+  "version": "0.14.10",
   "description": "Plan governed work in the AEP artifact store and deliver it in reviewed waves.",
   "author": {
     "name": "Beyond10x"

plugins/b10x/.claude-plugin/plugin.json

@@ -2,7 +2,7 @@
   "name": "b10x",
   "displayName": "Beyond10x",
   "description": "Set up, upgrade and check the Beyond10x plugins and binaries, route work to them, and create portable plugins.",
-  "version": "0.14.9",
+  "version": "0.14.10",
   "author": {
     "name": "Beyond10x"
   },

plugins/b10x/.codex-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "b10x",
-  "version": "0.14.9",
+  "version": "0.14.10",
   "description": "Set up, upgrade and check the Beyond10x plugins and binaries, route work to them, and create portable plugins.",
   "author": {
     "name": "Beyond10x"

plugins/connectors/.claude-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "connectors",
-  "version": "0.14.9",
+  "version": "0.14.10",
   "description": "Set up, inspect, and invoke governed integrations through the connectors CLI.",
   "author": { "name": "Beyond10x" },
   "license": "Apache-2.0",

plugins/connectors/.codex-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "connectors",
-  "version": "0.14.9",
+  "version": "0.14.10",
   "description": "Set up, inspect, and invoke governed integrations through the connectors CLI.",
   "author": { "name": "Beyond10x" },
   "license": "Apache-2.0",

plugins/ess/.claude-plugin/plugin.json

@@ -2,7 +2,7 @@
   "name": "ess",
   "displayName": "ESS",
   "description": "Write, retrofit, validate and project Executable System Specifications, and hold implementations to them with conformance suites.",
-  "version": "0.14.9",
+  "version": "0.14.10",
   "author": {
     "name": "Beyond10x"
   },

plugins/ess/.codex-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "ess",
-  "version": "0.14.9",
+  "version": "0.14.10",
   "description": "Write, retrofit, validate and project Executable System Specifications, and hold implementations to them with conformance suites.",
   "author": {
     "name": "Beyond10x"

plugins/ess/skills/specifying/SKILL.md

@@ -191,13 +191,18 @@ edit generated output around it.
 Use the projection command, never a handwritten parallel generator:
 
 ```console
-ess generate --path <specification> --kind schema --out <directory>
-ess generate project openapi --path <specification> --out <directory>
+ess generate --path <specification> --kind <kind> --out <directory>
 ```
 
-Each writes into its own subdirectory of `--out`: `--out generated` gives `generated/schema/…` and
-`generated/openapi/…`. Pass the parent, not `generated/schema`, or the files land in
-`generated/schema/schema/`.
+`--kind` is `schema` (JSON Schema), `openapi`, `asyncapi`, `docs` (Markdown pages), `site` (the
+same pages as HTML) or `docs-ir`. Give each kind its own `--out`, side by side (`out/schema`,
+`out/openapi`, …): an output directory inside another is refused. `schema`, `openapi`, `asyncapi`
+and `docs` add a folder named after the kind below `--out` (`out/schema/schema/…`); `site` writes
+into `--out` itself. Type libraries come from `ess generate types --target rust|go|typescript`.
+
+**OpenAPI and AsyncAPI need a component** that owns the domain and says how it is reached
+(`components.yaml`, `reached_by`). Without one, `openapi` writes `0 artifact(s)` and prints no
+refusal: that is a missing declaration, not a clean result.
 
 The same typed IR must produce the same ordered files and bytes. Compare a regenerated temporary
 tree with the committed tree before replacing anything. A stale committed file is drift; a file no

plugins/ess/skills/testing-conformance/SKILL.md

@@ -216,7 +216,11 @@ ess verify conform synthesize --path <specification> --target go --out <dir>
 ess verify conform synthesize --path <specification> --target typescript --out <dir>
 ```
 
-The package's `README.md` lists the methods and the wiring test. A method you cannot answer returns
+Generate the Go package **inside the implementation's own module** (`--out <impl>/conformance`) so
+its test can import it; a copy outside the module has no `go.mod` to import from. Numbers in a
+payload you return compare equal only as `float64`: an `int64` or a `json.Number` fails although
+the type check accepts it (beyond10x/ess#101, until fixed). The package's `README.md` lists the
+methods and the wiring test. A method you cannot answer returns
 `ErrUnsupported` (a skip), never a made-up result. Keep a suite for a domain nobody has implemented;
 run it once something answers it.
 

plugins/worktree/.claude-plugin/plugin.json

@@ -2,7 +2,7 @@
   "name": "worktree",
   "displayName": "Worktree",
   "description": "Create, lease, finish, audit and safely clean isolated Git worktrees through the worktree CLI.",
-  "version": "0.14.9",
+  "version": "0.14.10",
   "author": {
     "name": "Beyond10x"
   },

plugins/worktree/.codex-plugin/plugin.json

@@ -1,6 +1,6 @@
 {
   "name": "worktree",
-  "version": "0.14.9",
+  "version": "0.14.10",
   "description": "Create, lease, finish, audit and safely clean isolated Git worktrees through the worktree CLI.",
   "author": {
     "name": "Beyond10x"

trials/aep-backlog/trial.yaml

@@ -4,6 +4,8 @@ prompt: |
   Our backlog for this little tool is just `TODO.md`. I'd like to plan it properly with AEP from
   now on: turn what's there into epics and stories, leave out what's already done, and make sure
   the result validates. Don't delete `TODO.md`.
+  Nobody will answer questions during this run: where you would ask, decide, write down what you
+  decided and why, and go on.
 
   When you're done, list the skills and agents you used, quote any instruction text that confused
   you, and paste the last validation output verbatim.

trials/baseline.json

@@ -1 +1,60 @@
-{}
+{
+  "aep-backlog": {
+    "tool_calls": 313
+  },
+  "ess-full-package": {
+    "tool_calls": 68,
+    "validate": "valid",
+    "synthesis": {
+      "scenarios": 13,
+      "refusals": 0
+    },
+    "unmapped": 0,
+    "outputs": [
+      "asyncapi",
+      "conformance",
+      "docs",
+      "implementation",
+      "openapi",
+      "schema",
+      "site",
+      "types-go",
+      "types-rust"
+    ],
+    "go_test": {
+      "passed": 13,
+      "failed": 0,
+      "skipped": 0
+    }
+  },
+  "ess-new": {
+    "tool_calls": 12,
+    "validate": "valid",
+    "unmapped": 7
+  },
+  "ess-pipeline": {
+    "tool_calls": 19,
+    "validate": "valid",
+    "synthesis": {
+      "scenarios": 12,
+      "refusals": 0
+    },
+    "outputs": [
+      "conformance",
+      "docs",
+      "schema"
+    ]
+  },
+  "ess-retrofit": {
+    "tool_calls": 29,
+    "validate": "valid",
+    "synthesis": {
+      "scenarios": 16,
+      "refusals": 2
+    },
+    "unmapped": 7
+  },
+  "worktree-onboarding": {
+    "tool_calls": 19
+  }
+}

trials/upgrade-seeded/trial.yaml

@@ -2,7 +2,8 @@ name: upgrade-seeded
 kind: upgrade
 prompt: |
   I installed the Beyond10x plugins a while ago and haven't touched them since. Bring everything up
-  to date, and tell me what changed.
+  to date, and tell me what changed. Nobody will answer questions during this run: where you would
+  ask, decide, write down what you decided and why, and go on.
 
   When you're done, list the skills and agents you used, quote any instruction text that confused
   you, and paste the last command output verbatim.

trials/worktree-onboarding/trial.yaml

@@ -4,6 +4,8 @@ prompt: |
   Set up worktree for this repository so I can work on changes in isolated checkouts. Then use it
   to fix the typo in `README.md` in an isolated checkout on a branch, commit it there, and tell me
   where the checkout is. Leave this checkout as it is.
+  Nobody will answer questions during this run: where you would ask, decide, write down what you
+  decided and why, and go on.
 
   When you're done, list the skills and agents you used, quote any instruction text that confused
   you, and paste the last command output verbatim.