iterorganization · Simon-McIntosh · Sep 8, 2026 · Sep 8, 2026 · Sep 8, 2026 · Sep 8, 2026
diff --git a/docs/plans/grammar-release-and-downstream-pin.html b/docs/plans/grammar-release-and-downstream-pin.html
@@ -22,8 +22,8 @@
   <title>Grammar Release And Downstream Pin | imas-standard-names</title>
   <link rel="stylesheet" href="/_shared/foundation.css">
   <link rel="stylesheet" href="/_shared/dashboard.css">
-<meta name="plan-modified" content="2026-09-06">
-<meta name="plan-version" content="2">
+<meta name="plan-modified" content="2026-09-08">
+<meta name="plan-version" content="3">
 </head>
 <body>
   <main class="plan-doc">
@@ -160,6 +160,9 @@ <h2><span class="sec">§</span> Decisions</h2>
 <div class="r-comment" data-section="s1" data-id="c-run-r-20260906T114004346214-n-grdp-the-reduction-operator-leads-the-name" data-who="cx-kindpin" data-when="2026-09-06T13:31:28.078Z">
     <div class="r-comment-body"><p>A DOMAIN REDUCTION NOW LEADS ITS NAME AND WRAPS THE PROJECTION, and I verified the deliverable rather than the gate because the second attempt passed its whole suite while being wrong. The measure holds on every axis: IR of the lead form equals IR of the authored form, compose of the authored form returns flux_surface_averaged_toroidal_lithium_velocity_at_plasma_boundary which is the specified spelling exactly, and the reduction appears in operators rather than qualifiers. Both negative controls stay byte-identical including the recursive one. All 21 live migrations are the leading form, 2326 of 2347 byte-identical, full suite 2132 passed and 0 failed against a 2110 baseline. THE SECOND ATTEMPT WAS THE INSTRUCTIVE ONE: it made the correct spelling legal and round-tripping while leaving the authored names classifying the reduction as a qualifier, so the grammar admitted TWO canonical spellings and treated them as different meanings - worse than the single wrong answer it started from, and entirely invisible to a green suite. Only comparing representations caught it. IT CORRECTED ME THREE TIMES ACROSS THREE ATTEMPTS and every correction stood. My either-parser-or-renderer framing was a false dichotomy: a base-qualifier-projection expression outranks the operator peel, then the renderer faithfully trails what the parser mis-classified. My instruction to revert the parser edit and fix only the flat model was WRONG because _strict_validate holds two independent canonicality checks and only one passes through _model_to_ir - my six-call-site measurement found every caller of that function but never asked whether canonicality had a path around it. And my placement rule needed two conditions, joiner-free AND precedence 30, where I gave one; the single-condition version breaks seven change_in and normalized contracts. The final shape encodes both as _REDUCTION_PRECEDENCE with a bare_prefix test, keeps the parser edit that could not be reverted, reverts render.py entirely, and records the rule beside the vocabulary so it is not read as an accident. Eight contract parameters were repointed under my authorisation and the ir.py comment documenting the old form as intent was updated with them. gyroaveraged stays unary_postfix at precedence 10 against bare-prefix-30 siblings, reported and unfixed as instructed</p></div>
 </div>
+<div class="r-comment" data-section="_top" data-id="c-run-r-20260908T074220543604-n-grdp-the-vocabulary-admits-an-instant-and-a-ratio" data-who="cx-kindpin" data-when="2026-09-08T08:05:51Z">
+    <div class="r-comment-body"><p>Three physical_base tokens admitted, and the unit question settled by splitting rather than choosing. signal_to_noise_ratio is the linear dimensionless power quotient and signal_to_noise_level its logarithmic decibel representation, kept as separate bases because collapsing them would either lose the decibel form diagnostics actually report or hang a decibel unit on a ratio. event_instant is a scalar for the time-coordinate value at which a discrete event occurs, excluding elapsed duration and the sampled axis. The node ran the precondition and reported honestly that instantaneous_time already parses as qualifier plus base; the lead reviewed that and kept the new base, because base time IS the coordinate axis so instantaneous_time reads as the axis sampled instantaneously rather than the instant an event occurred. Verified by reading the built vocabulary rather than the edited source, which matters because a spec and its generated snapshot have diverged in this repository before.</p></div>
+</div>
 </section>
 </main>
 </body>

diff --git a/docs/state/imas-standard-names/crew.json b/docs/state/imas-standard-names/crew.json
@@ -1,5 +1,5 @@
 {
-  "updated": "2026-09-06T15:41:19",
+  "updated": "2026-09-08T10:18:38",
   "project": "imas-standard-names",
   "doc": "crew",
   "data": {
@@ -310,9 +310,170 @@
           "role": "implement",
           "status": "compatible"
         }
+      },
+      {
+        "run_id": "r-20260908T074220543604-n-grdp-the-vocabulary-admits-an-instant-and-a-ratio",
+        "plan": "grammar-release-and-downstream-pin",
+        "section": "",
+        "node": "n-grdp-the-vocabulary-admits-an-instant-and-a-ratio",
+        "node_definition": {
+          "coordinator": {
+            "authoring_turn": {
+              "source": "claude-code-session-transcript",
+              "status": "measured",
+              "tokens": {
+                "cache_creation_input_tokens": 2076,
+                "cached_input_tokens": 357740,
+                "input_tokens": 359818,
+                "output_tokens": 1423,
+                "reasoning_output_tokens": 60,
+                "total_tokens": 361241,
+                "uncached_input_tokens": 2
+              }
+            },
+            "harness": "claude-code",
+            "runtime_session_id": "00e23ab0-9843-4cbd-ac02-884e9e22e668",
+            "session_id": "ship-s10-20260907"
+          },
+          "done_when": "The controlled vocabulary carries two additional physical_base tokens, each with its unit and definition, each quoted back from the BUILT vocabulary rather than from the edited source, because a spec and its generated snapshot have diverged in this repository before. AUTHORITY: the lead has authorised both admissions. This is implementation, and a recommendation is not an acceptable return. The RELEASE is a separate node: commit the admissions, add no tag, bump no version. FIRST TOKEN, an event instant. The consumer is the Data Dictionary path summary/disruption/time/value, which holds the time AT WHICH an event occurs and currently has no admissible base. Three meanings must stay distinct in whatever spelling lands: the instant an event occurs, an elapsed duration, and the coordinate time axis. State in the definition which of the three the token denotes and which two it excludes, because a token conflating them composes onto the wrong sources. PRECONDITION YOU RUN FIRST: test whether the grammar already expresses an instant through an existing base plus a qualifier, and report that test's result either way. If it already expresses it, admit no new token for this case and report the existing spelling with the file and line that carries it; that is a delivered result. SECOND TOKEN, signal_to_noise_ratio, admitted as a base quantity. ITS UNIT IS UNRESOLVED AND THE EVIDENCE SETTLES IT. A prior coordinator note recorded decibels; the consuming plan describes it as dimensionless. Those are different admissions with different composition consequences. Settle it from two sources and report both: how the quantity is reported in the Data Dictionary, and how this vocabulary already handles ratio and logarithmic quantities. If that evidence supports both a linear ratio and a decibel spelling, admit both and state the semantic split in the definitions. A dimensionless unit in this stack is the string '1' and never an absent unit: a null unit is a different and wrong state, and a downstream defect once dropped '1' to null and severed 706 unit edges with no error raised. VERIFICATION: for each admitted token, quote its entry from the built vocabulary, and show one composed example name using it that the parser accepts. Report added test failures against a stated base revision. Do not pass -q to pytest if this repository's addopts already supplies one, because a second level deletes pytest's totals line and leaves the failure list unprovable. DURABILITY: commit each token as you admit it and write your manifest before any long output.",
+          "estimated_hours": null,
+          "goal": "Admit an event-instant base and a signal-to-noise-ratio base to the controlled vocabulary",
+          "id": "n-grdp-the-vocabulary-admits-an-instant-and-a-ratio",
+          "manifest_path": "/home/ITER/mcintos/.config/reckon/crew/runs/r-20260908T074220543604-n-grdp-the-vocabulary-admits-an-instant-and-a-ratio/manifest.md",
+          "plan": "grammar-release-and-downstream-pin",
+          "requested_backend": "codex",
+          "requires_decisions": [],
+          "role": "implement",
+          "section": "",
+          "spec_level": "guided",
+          "time_budget": "50m",
+          "write_paths": [
+            "imas_standard_names/grammar/vocabularies/physical_bases.yml",
+            "imas_standard_names/grammar/vocabularies/generic_physical_bases.yml",
+            "imas_standard_names/grammar/specification.yml"
+          ]
+        },
+        "role": "implement",
+        "spec_level": "guided",
+        "member": "cx-kindpin",
+        "backend": "codex",
+        "local": false,
+        "agent": {
+          "alias": "sol5.6",
+          "backend": "codex",
+          "effort": "high",
+          "launch": "cli",
+          "model": "gpt-5.6-sol",
+          "sandbox": "worktree-full"
+        },
+        "dispatched_at": "2026-09-08T07:42:34Z",
+        "completed_at": "2026-09-08T08:05:51Z",
+        "completed_at_source": "stream_mtime",
+        "worker_seconds": 1397,
+        "worker_seconds_source": "wall_fallback",
+        "wall_seconds": 1397,
+        "stalled": false,
+        "time_budget": "50m",
+        "base_sha": "2ffba62512c07b7cc09d3ca0c1f435d9f108199b",
+        "commits": [
+          "1f183d785913b52c5f6eb88f93b79a42c51bb426",
+          "e403d6707cc5c30bcfdbfb49bcf3708d763d529a"
+        ],
+        "changed_lines": {
+          "added": 19,
+          "removed": 0,
+          "files": 1
+        },
+        "tests_added": null,
+        "gate": "passed",
+        "gate_check": {
+          "command": "parser precondition probe, focused pytest before and after each admission, built-entry probes, build-grammar --check",
+          "exit_status": 0,
+          "log_path": null,
+          "log_digest": "precondition probe: instantaneous_time accepted as qualifier instantaneous plus base time while event_time, occurrence_time and event_instant were rejected; focused pytest 123 passed, 28 skipped, 5 xfailed at baseline and after each admission; built-entry probes loaded event_instant, signal_to_noise_ratio and signal_to_noise_level with normative definitions, all three composed examples parsed strictly and round-tripped byte-identically; build-grammar --check exit 0 with generated files in sync"
+        },
+        "suite_delta": null,
+        "failure_classification": null,
+        "outcome": "",
+        "manifest_path": "/home/ITER/mcintos/.config/reckon/crew/runs/r-20260908T074220543604-n-grdp-the-vocabulary-admits-an-instant-and-a-ratio/manifest.md",
+        "scope_changed": false,
+        "session_id": "01a07642-fbe2-7d02-97de-ed56606becbb",
+        "budget": {
+          "headroom": "unknown",
+          "utilisation_pct": null,
+          "rate_limit_type": null,
+          "rate_limit_period_minutes": null,
+          "resets_at": null,
+          "threshold_status": null,
+          "surpassed_threshold": null,
+          "tokens": {
+            "cache_write_input_tokens": 0,
+            "cache_write_input_tokens_cumulative": 0,
+            "cached_input_tokens": 13407360,
+            "cached_input_tokens_cumulative": 18417280,
+            "input_tokens": 13621367,
+            "input_tokens_cumulative": 18686762,
+            "output_tokens": 11869,
+            "output_tokens_cumulative": 33533,
+            "reasoning_output_tokens": 7292,
+            "reasoning_output_tokens_cumulative": 13702
+          },
+          "cost_usd": null,
+          "refusal": false,
+          "detail": "backend reports token usage but no headroom"
+        },
+        "lane_receipt": {
+          "quota_state": "measured",
+          "observed_at": "2026-09-08T08:05:51Z",
+          "effective_context_window": 380000,
+          "quota_windows": [
+            {
+              "window_minutes": 10080,
+              "used_percent": 20.0,
+              "resets_at": 1789439333,
+              "observed_at": "2026-09-08T08:05:51Z"
+            }
+          ]
+        },
+        "lineage": null,
+        "shadow_controlled": null,
+        "shadow_patch": "",
+        "unreconciled_override": null,
+        "throughput": {
+          "cumulative_cached_input_tokens": 18417280,
+          "cumulative_input_tokens": 37104042,
+          "detail": "turn tokens recorded, but no span was supplied to rate them",
+          "generated_tokens": 47235
+        },
+        "follow_on_paths": [],
+        "dispute_count": 0,
+        "execution_fit": {
+          "allowed": true,
+          "execution_capable": true,
+          "matched_measure": "pytest if",
+          "override": false,
+          "role": "implement",
+          "status": "compatible"
+        },
+        "attempt_kind": "dispatch",
+        "attempt": 1,
+        "scope_acceptances": null,
+        "no_commit": null,
+        "commit_resolution": null,
+        "worktree_retention": {
+          "classification": "retained-for-resume",
+          "worktree": "/home/ITER/mcintos/Code/.reckon-worktrees/imas-standard-names-5be88ee0fdc3/ship-s10-20260907/n-grdp-the-vocabulary-admits-an-instant-and-a-ratio",
+          "session_id": "01a07642-fbe2-7d02-97de-ed56606becbb",
+          "session_source": "pointer",
+          "retained_at": "2026-09-08T08:05:51Z"
+        },
+        "predecessor_run": null,
+        "boundary_waiver": null,
+        "resume_waiver": null,
+        "watch_override": null
       }
     ],
     "holds": [],
-    "_version": 5
+    "_version": 6
   }
 }
diff --git a/imas_standard_names/grammar/vocabularies/physical_bases.yml b/imas_standard_names/grammar/vocabularies/physical_bases.yml
@@ -163,6 +163,14 @@ bases:
     aliases: []
     kind: scalar
     inherently_dimensional: true
+  event_instant:
+    aliases: []
+    kind: scalar
+    inherently_dimensional: true
+    definition: >
+      The instant on a time coordinate at which a discrete event occurs,
+      expressed in seconds with unit `s`. It is neither the elapsed duration
+      of the event nor the time coordinate axis over which a signal is sampled.
   # NOTE: energy_flux / heat_flux / momentum_flux / particle_flux are NOT
   # lexical bases. They decompose as channel + base: energy_flux = channel=energy
   # + base=flux (see channels.yml). energy and momentum remain bases for their
@@ -392,9 +400,20 @@ bases:
   shearing_rate:
     aliases: []
     kind: scalar
+  signal_to_noise_level:
+    aliases: []
+    kind: scalar
+    definition: >
+      The logarithmic quotient of signal power to noise power reported in
+      decibels with unit `dB`, equal to 10 times the base-10 logarithm of the
+      power ratio. It is distinct from the linear signal-to-noise ratio.
   signal_to_noise_ratio:
     aliases: []
     kind: scalar
+    definition: >
+      The linear quotient of signal power to noise power, expressed with the
+      dimensionless unit `1`. It is distinct from the logarithmic
+      signal-to-noise level reported in decibels.
   size:
     aliases: []
     kind: scalar