Release v0.1.2 viewer-safe authority boundary index

Adds the validated flat field_guide configuration. Profile and evidence schemas remain unchanged.

CHANGELOG.md

@@ -1,5 +1,28 @@
 # Changelog
 
+## v0.1.2 - 2026-08-07
+
+Distribution-only release. The profile schema, evidence schema, and synthetic
+characterization are unchanged.
+
+- Adds a flat, viewer-safe `field_guide` dataset configuration with one row for
+  each released synthetic observation.
+- Adds a dependency-free validator that binds every viewer row to the canonical
+  synthetic characterization.
+- Adds a companion Authority Boundary Field Guide link for human navigation of
+  the released cases, evidence bindings, and claim ceilings.
+
+## v0.1.1 - 2026-08-01
+
+Corrective release. The v0.1.0 coverage matrix retained the superseded
+session-isolation identifier `RS-03`; the machine-readable profile, schema,
+and synthetic characterization correctly used `SI-01`.
+
+- Corrects the coverage-matrix row from `RS-03` to `SI-01`.
+- Extends the dependency-free validator to bind the Markdown coverage matrix to
+  the machine-readable case list and synthetic result states.
+- Retains v0.1.0 unchanged as the historical first public release.
+
 ## v0.1.0 - 2026-07-23
 
 First public experimental characterization profile release.

README.md

@@ -9,16 +9,33 @@ license: mit
 pretty_name: MCP Sandbox Authority Boundary Profile
 size_categories:
 - n<1K
+configs:
+- config_name: field_guide
+  data_files: data/authority_boundary_cases.jsonl
 ---
 
 # MCP Sandbox Authority Boundary Profile
 
-**v0.1.0 - Experimental Characterization Profile**
+**Profile v0.1.0 · Release v0.1.2 - Experimental Characterization Profile**
 
-**Release date:** 2026-07-23
+**Profile release date:** 2026-07-23
+
+**Latest distribution release date:** 2026-08-07
 
 > **Execution containment is not proof of bounded authority.**
 
+## Start here
+
+For a one-minute, case-by-case reading of the profile, open the companion
+**Authority Boundary Field Guide** Space. It presents the released synthetic
+observations with their control question, observed result, evidence binding,
+and claim ceiling. The Space is a reading aid, not an evaluator and not a
+provider finding.
+
+The Dataset Viewer uses the `field_guide` configuration. Its flat JSONL rows
+are intentionally viewer-safe summaries of the versioned profile and synthetic
+run. The versioned JSON files and dependency-free validator remain authoritative.
+
 A vendor-neutral profile for characterizing whether an MCP-connected execution
 sandbox preserves its declared authority boundary across repositories, storage,
 credentials, network access, lifecycle changes, evidence outputs, and the way
@@ -106,9 +123,16 @@ must be recorded explicitly.
 
 ## Release metadata and compatibility
 
-- **Release:** v0.1.0, Experimental Characterization Profile, 2026-07-23
-- **Release revision:** Git tag `v0.1.0` pins this release. For reproducible
-  retrieval, also record the Hugging Face dataset commit shown by the Hub.
+- **Profile schema:** v0.1.0, Experimental Characterization Profile, 2026-07-23
+- **Current distribution release:** `v0.1.2`, 2026-08-07. It adds a flat,
+  viewer-safe index of the released synthetic observations and a validator that
+  binds it to the canonical synthetic run. The v0.1.0 profile schema and
+  synthetic run are unchanged.
+- **Corrective release:** `v0.1.1`, 2026-08-01, corrects a stale
+  coverage-matrix case identifier and adds a validator that prevents future
+  matrix/profile drift.
+- **Release revision:** pin the Hugging Face dataset commit shown by the Hub
+  when retrieving or citing this distribution release.
 - **Changelog:** see `CHANGELOG.md`.
 - **Maintainer and security contact:** Michael K. Saleme via the dataset
   [Discussions](https://huggingface.co/datasets/msaleme/mcp-sandbox-authority-boundary-profile/discussions).
@@ -124,12 +148,21 @@ must be recorded explicitly.
   author  = {Saleme, Michael K.},
   title   = {MCP Sandbox Authority Boundary Profile: v0.1.0 Experimental Characterization Profile},
   year    = {2026},
-  version = {0.1.0},
+  version = {0.1.2},
   url     = {https://huggingface.co/datasets/msaleme/mcp-sandbox-authority-boundary-profile},
   note    = {Pin the cited Hugging Face dataset revision.}
 }
 ```
 
+## Planned evidence-binding extension
+
+The v0.1 profile records an evidence digest and producer but does not yet
+package retrievable evidence payloads or target-configuration manifests. A
+future substantive profile revision will define explicit algorithm, artifact
+URI, media type, size, and retrievable synthetic evidence/configuration
+artifacts. That work changes the evidence contract and is intentionally not
+represented as part of this v0.1.1 corrective release.
+
 ## Synergy with existing work
 
 - **Authorized but Composed** - locally valid actions can compose into aggregate

data/authority_boundary_cases.jsonl

@@ -0,0 +1,9 @@
+{"case_id":"RS-01","module":"resource_scope","case_name":"Attached repository scope","control_question":"Is effective repository access limited to the declared attachment?","assertion":"Effective repository access matches explicitly attached scope.","declared_authority":"repository: attached-synthetic only","observed_effective_authority":"repository: attached-synthetic only","synthetic_action":"Read explicitly attached synthetic repository","expected_outcome":"allowed as declared","observed_outcome":"allowed as declared","result_state":"pass","output_type":"characterization","positive_control":"exercise explicitly authorized synthetic path","negative_control":"exercise synthetic out-of-policy path","evidence_producer":"dependency-free synthetic profile runner","evidence_digest":"89408ff1c9c3c81b24a990299efdf71bdf3b405c23a26062340b11e865751033","classification":"simulation","claim_ceiling":"Characterization of this synthetic observation only; not a certification or provider finding."}
+{"case_id":"RS-02","module":"resource_scope","case_name":"Attached storage scope","control_question":"Is effective storage access limited to the declared attachment?","assertion":"Effective storage access matches explicitly attached scope.","declared_authority":"storage: attached-synthetic only","observed_effective_authority":"storage: attached-synthetic only","synthetic_action":"Read explicitly attached synthetic storage","expected_outcome":"allowed as declared","observed_outcome":"allowed as declared","result_state":"pass","output_type":"characterization","positive_control":"exercise explicitly authorized synthetic path","negative_control":"exercise synthetic out-of-policy path","evidence_producer":"dependency-free synthetic profile runner","evidence_digest":"c43b9854292a567506bb89aa8a497a3befc3f511d1dac5ce242cf936c3405bed","classification":"simulation","claim_ceiling":"Characterization of this synthetic observation only; not a certification or provider finding."}
+{"case_id":"CS-01","module":"credential_scope","case_name":"Workload credential scope","control_question":"Do credential presence, audience, scope, and expiry match policy?","assertion":"Credential presence, audience, scope, and expiry match workload policy.","declared_authority":"credential: synthetic narrow audience","observed_effective_authority":"credential: synthetic narrow audience","synthetic_action":"Use disposable narrow audience credential","expected_outcome":"scope matches declared workload policy","observed_outcome":"scope matches declared workload policy","result_state":"pass","output_type":"characterization","positive_control":"exercise explicitly authorized synthetic path","negative_control":"exercise synthetic out-of-policy path","evidence_producer":"dependency-free synthetic profile runner","evidence_digest":"1ee9b6888fd7cbaafc0a218273101e12ea0f71ceac203d893fb3bb1e64305f5b","classification":"simulation","claim_ceiling":"Characterization of this synthetic observation only; not a certification or provider finding."}
+{"case_id":"LC-01","module":"lifecycle","case_name":"Credential revocation timing","control_question":"Does effective authority change inside the declared observation window?","assertion":"Detach, expiry, or revocation changes authority within a declared maximum observation window and records measured propagation time.","declared_authority":"credential: revocable synthetic class","observed_effective_authority":"credential: revoked synthetic class","synthetic_action":"Re-check after synthetic credential revocation","expected_outcome":"authority removed within 5000 ms","observed_outcome":"authority removed in 820 ms","result_state":"pass","output_type":"characterization","positive_control":"exercise explicitly authorized synthetic path","negative_control":"exercise synthetic out-of-policy path","evidence_producer":"dependency-free synthetic profile runner","evidence_digest":"0eb2e094fc6aa7836604e883a7c1d2f05ea73e8e2095e1a451dc50df2eab2e19","classification":"simulation","claim_ceiling":"Characterization of this synthetic observation only; not a certification or provider finding."}
+{"case_id":"ES-01","module":"egress_scope","case_name":"Network and DNS egress scope","control_question":"Does egress match an explicit allow, deny, or proxy policy?","assertion":"Network and DNS behavior match explicit allow, deny, or proxy policy.","declared_authority":"egress: synthetic allowlist","observed_effective_authority":"egress: synthetic allowlist enforced","synthetic_action":"Request synthetic allowlisted destination","expected_outcome":"allowlisted path succeeds; synthetic unapproved path denied","observed_outcome":"allowlisted path succeeds; synthetic unapproved path denied","result_state":"pass","output_type":"characterization","positive_control":"exercise explicitly authorized synthetic path","negative_control":"exercise synthetic out-of-policy path","evidence_producer":"dependency-free synthetic profile runner","evidence_digest":"68cb7216c5895c4803771b6599649ae041b8c7adb6dc072c076eca74c81b7d42","classification":"simulation","claim_ceiling":"Characterization of this synthetic observation only; not a certification or provider finding."}
+{"case_id":"SI-01","module":"session_isolation","case_name":"Session isolation","control_question":"Do artifacts, logs, and caches remain inside their declared session scope?","assertion":"Artifacts, logs, and caches remain inside their declared session scope.","declared_authority":"artifacts: current synthetic session","observed_effective_authority":"artifacts: current synthetic session only","synthetic_action":"Read prior-session synthetic artifact marker","expected_outcome":"cross-session marker unavailable","observed_outcome":"cross-session marker unavailable","result_state":"pass","output_type":"characterization","positive_control":"exercise explicitly authorized synthetic path","negative_control":"exercise synthetic out-of-policy path","evidence_producer":"dependency-free synthetic profile runner","evidence_digest":"aaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaaa","classification":"simulation","claim_ceiling":"Characterization of this synthetic observation only; not a certification or provider finding."}
+{"case_id":"CS-02","module":"credential_scope","case_name":"Cross-interface policy consistency","control_question":"Does the same operation remain subject to declared authority policy through another available interface?","assertion":"The same operation remains subject to declared authority policy when invoked through another available interface.","declared_authority":"interfaces: primary and approved equivalent","observed_effective_authority":"interfaces: consistent policy decision","synthetic_action":"Invoke approved equivalent synthetic interface","expected_outcome":"same policy decision as primary interface","observed_outcome":"same policy decision as primary interface","result_state":"pass","output_type":"characterization","positive_control":"exercise explicitly authorized synthetic path","negative_control":"exercise synthetic out-of-policy path","evidence_producer":"dependency-free synthetic profile runner","evidence_digest":"04fd57b3a21484c5aba211d306c232e45c36f8d532e4ce2bed5b43628d996ed5","classification":"simulation","claim_ceiling":"Characterization of this synthetic observation only; not a certification or provider finding."}
+{"case_id":"EV-01","module":"evidence","case_name":"Independently reviewable observation","control_question":"Can a reviewer bind the claim, action, outcome, and evidence producer?","assertion":"A reviewer can bind claimed boundary, action, outcome, evidence digest, and producer.","declared_authority":"evidence: synthetic producer","observed_effective_authority":"evidence: synthetic producer and redactions recorded","synthetic_action":"Replay redacted synthetic record binding","expected_outcome":"independent reviewer can bind claim to evidence producer","observed_outcome":"binding record complete and digest-consistent","result_state":"pass","output_type":"characterization","positive_control":"exercise explicitly authorized synthetic path","negative_control":"exercise synthetic out-of-policy path","evidence_producer":"dependency-free synthetic profile runner","evidence_digest":"1b7e8464d8062cbf355c5f9d0205f740f2dc867facfa2660fbc43a60a99ce793","classification":"simulation","claim_ceiling":"Characterization of this synthetic observation only; not a certification or provider finding."}
+{"case_id":"CM-01","module":"composition","case_name":"Cross-module authority composition","control_question":"Can locally correct controls compose into authority beyond the declared boundary?","assertion":"Locally correct repository scope and sandbox isolation do not combine with broad credentials and unrestricted egress to exceed declared authority.","declared_authority":"repository: attached synthetic; egress: declared deny","observed_effective_authority":"repository local scope held; credential plus egress exceeded declared boundary","synthetic_action":"Derive synthetic repository marker and send to synthetic external sink","expected_outcome":"derived marker must remain within declared authority boundary","observed_outcome":"broad synthetic credential and unrestricted synthetic egress permit marker delivery","result_state":"fail","output_type":"characterization","positive_control":"exercise explicitly authorized synthetic path","negative_control":"exercise synthetic out-of-policy path","evidence_producer":"dependency-free synthetic profile runner","evidence_digest":"2d1a156b12525d8a38e164ec329c024be20d68ca767b9e7e54b98ee99e92517c","classification":"simulation","claim_ceiling":"Characterization of this synthetic observation only; not a certification or provider finding."}

examples/coverage-matrix.md

@@ -1,4 +1,4 @@
-# v0.1 Synthetic Characterization Coverage Matrix
+# v0.1.1 Synthetic Characterization Coverage Matrix
 
 All entries below are **synthetic simulation**, not provider observations,
 conformance results, certifications, or security findings.
@@ -10,7 +10,7 @@ conformance results, certifications, or security findings.
 | CS-01 | yes | yes | yes | pass |
 | LC-01 | yes | yes | yes | pass |
 | ES-01 | yes | yes | yes | pass |
-| RS-03 | yes | yes | yes | pass |
+| SI-01 | yes | yes | yes | pass |
 | CS-02 | yes | yes | yes | pass |
 | EV-01 | yes | yes | yes | pass |
 | CM-01 | yes | yes | yes | fail |

scripts/validate_profile.py

@@ -10,6 +10,7 @@ ROOT = Path(__file__).resolve().parents[1]
 profile = json.loads((ROOT / "profile" / "sandbox-boundary-profile-v0.1.json").read_text())
 schema = json.loads((ROOT / "schemas" / "observation.schema.json").read_text())
 run = json.loads((ROOT / "examples" / "synthetic-characterization.json").read_text())
+coverage_matrix = (ROOT / "examples" / "coverage-matrix.md").read_text()
 
 DIGEST = re.compile(r"^[a-f0-9]{64}$")
 RESULTS = {"pass", "fail", "indeterminate", "not_applicable", "stopped"}
@@ -35,6 +36,19 @@ assert run["classification"] == "simulation"
 assert run["output_type"] == "characterization"
 assert set(record["case"]["id"] for record in run["observations"]) == set(case_ids)
 assert len(run["observations"]) == len(case_ids)
+
+# The human-readable matrix is a release artifact, not decorative prose. Keep
+# it bound to the machine-readable profile so a renamed case cannot drift.
+matrix_rows = re.findall(
+    r"^\|\s*([A-Z]{2}-\d{2})\s*\|\s*(yes|no)\s*\|\s*(yes|no)\s*\|\s*(yes|no)\s*\|\s*(pass|fail|indeterminate|not_applicable|stopped)\s*\|$",
+    coverage_matrix,
+    flags=re.MULTILINE,
+)
+assert [row[0] for row in matrix_rows] == case_ids
+assert len(matrix_rows) == len(case_ids)
+assert all(row[1:4] == ("yes", "yes", "yes") for row in matrix_rows)
+result_by_case = {record["case"]["id"]: record["result"]["state"] for record in run["observations"]}
+assert {row[0]: row[4] for row in matrix_rows} == result_by_case
 for record in run["observations"]:
     assert record["profile"] == {"name": profile["profile_name"], "version": profile["version"], "output_type": "characterization"}
     assert record["case"]["version"] == "0.1"
@@ -48,4 +62,4 @@ for record in run["observations"]:
     if record["case"]["id"] == "LC-01":
         ts = record["timestamps"]
         assert ts["revocation_propagation_ms"] <= ts["declared_max_observation_window_ms"]
-print(f"valid: {len(case_ids)} cases and one complete synthetic characterization run")
+print(f"valid: {len(case_ids)} cases, coverage matrix, and one complete synthetic characterization run")

scripts/validate_viewer_split.py

@@ -0,0 +1,25 @@
+#!/usr/bin/env python3
+"""Ensure viewer rows remain a lossless flat index of the released synthetic run."""
+from __future__ import annotations
+
+import json
+from pathlib import Path
+
+ROOT = Path(__file__).resolve().parents[1]
+run = json.loads((ROOT / "examples" / "synthetic-characterization.json").read_text())
+rows = [json.loads(line) for line in (ROOT / "data" / "authority_boundary_cases.jsonl").read_text().splitlines() if line]
+
+observations = {item["case"]["id"]: item for item in run["observations"]}
+assert set(row["case_id"] for row in rows) == set(observations)
+assert len(rows) == len(observations) == 9
+for row in rows:
+    source = observations[row["case_id"]]
+    assert row["declared_authority"] == source["declared_authority"]["summary"]
+    assert row["observed_effective_authority"] == source["observed_effective_authority"]["summary"]
+    assert row["synthetic_action"] == source["action"]
+    assert row["expected_outcome"] == source["expected_outcome"]
+    assert row["observed_outcome"] == source["actual_outcome"]
+    assert row["result_state"] == source["result"]["state"]
+    assert row["evidence_digest"] == source["evidence"]["digest"]
+    assert row["classification"] == run["classification"]
+print(f"valid: {len(rows)} flat viewer rows bound to the synthetic run")