release: v0.6.0

Adopting-this-into-your-project README guide (six-step onboarding path),
plus the carried-forward v0.6-prep fix from the private staging repo:
agent-permission-guard.py FORBIDDEN_PATTERNS regex hardened to be
case-insensitive and whitespace-tolerant, with a pinning regression test.

See CHANGELOG.md [0.6.0] for the full entry.

Author: onesimplecode

CHANGELOG.md

@@ -5,6 +5,15 @@ Format follows [Keep a Changelog](https://keepachangelog.com/en/1.1.0/).
 
 ## [Unreleased]
 
+## [0.6.0] - 2026-07-20
+
+### Added
+
+- `README.md` — "Adopting this into your project" section: a six-step path
+  (operating model → `AGENTS.md`/`agents/` → scripts against your own repo →
+  templates → worked examples → `agent-skills-integration.md`) consolidating
+  guidance that was previously scattered across the README and `docs/`
+
 ## [0.5.0] - 2026-07-16
 
 ### Added
@@ -223,7 +232,8 @@ Format follows [Keep a Changelog](https://keepachangelog.com/en/1.1.0/).
 - `CONTRIBUTING.md`, `SECURITY.md`, issue/PR templates, `release-check` CI workflow
 - Roadmap and changelog for intentional release cadence
 
-[Unreleased]: https://github.com/onesimplecode/ai-engineering-standards/compare/v0.5.0...HEAD
+[Unreleased]: https://github.com/onesimplecode/ai-engineering-standards/compare/v0.6.0...HEAD
+[0.6.0]: https://github.com/onesimplecode/ai-engineering-standards/compare/v0.5.0...v0.6.0
 [0.5.0]: https://github.com/onesimplecode/ai-engineering-standards/compare/v0.4.0...v0.5.0
 [0.4.0]: https://github.com/onesimplecode/ai-engineering-standards/compare/v0.3.0...v0.4.0
 [0.3.0]: https://github.com/onesimplecode/ai-engineering-standards/compare/v0.2.0...v0.3.0

README.md

@@ -42,6 +42,34 @@ python3 scripts/llms-txt-generator.py
 python3 scripts/agent-permission-guard.py --settings /path/to/your/settings.json
 ```
 
+## Adopting this into your project
+
+A step-by-step path for pulling these standards into your own repo, not just this one:
+
+1. **Read the model** — [`docs/ai-engineering-operating-model.md`](docs/ai-engineering-operating-model.md)
+   explains the four layers (requirements, roles, artifacts, checks) and the
+   failure modes they guard against.
+2. **Pull in agent conventions** — copy or reference [`AGENTS.md`](AGENTS.md) and
+   the role specs in [`agents/`](agents/) into your own repo so any AI coding
+   tool reads the same rules.
+3. **Run the checks against your repo** — point the Quick start scripts above at
+   your own monorepo (`--root /path/to/your/repo`) instead of this one.
+4. **Adopt templates as needed** — the ADR, impact assessment, maturity
+   checklist, LLM eval, and completion checklist templates in
+   [`templates/`](templates/) are meant to be copied, not just read.
+5. **Study the worked traces** — [`examples/worked-example/`](examples/worked-example/)
+   and [`examples/agent-permission-guard/`](examples/agent-permission-guard/) show
+   a requirement moving end-to-end: TR-ID → ADR → maturity row → script → CI gate.
+6. **Reconcile with tools you already use** — [`docs/agent-skills-integration.md`](docs/agent-skills-integration.md)
+   covers how this layers under AGENTS.md, agent-skills, and Cursor rules rather
+   than replacing them.
+
+None of this requires forking the repo — steps 2–4 are copy-in, and every
+script in step 3 takes a root-path argument (`--root`, `--path`, `--out`,
+`--settings`, or a positional path, depending on the script — see each
+command's `--help`) precisely so it can target your own repo instead of this
+one.
+
 ## Enforced workflow
 
 See [`examples/worked-example/`](examples/worked-example/) for a synthetic trace:

ROADMAP.md

@@ -84,10 +84,18 @@ See `CHANGELOG.md` for shipped versions.
       (v0.4 item 1) to any agent framework that reads the emerging llms.txt
       convention (https://llmstxt.org), not just Cursor
 
-## v0.6 — Agentic security & operations patterns
+## v0.6 — Adoption guide
+
+- [x] `README.md` "Adopting this into your project" section: a six-step path
+      (operating-model doc, `AGENTS.md`/`agents/`, running the Quick-start
+      scripts against the reader's own repo, `templates/`, the two worked
+      examples, `docs/agent-skills-integration.md`) consolidating onboarding
+      guidance that was previously scattered across the README and `docs/`
+
+## v0.7 — Agentic security & operations patterns (current)
 
 From the 2026-07-13 Zero-Trust-for-AI-Agents review (private ADRs: private-repo
-ADR-030/031/032, private-repo ADR-018, cloud deployment proposal rev 7). Two
+ADR-030/031/032, private-repo ADR-018, and a private-repo deployment proposal). Two
 maturity classes — this repo exports packaged practice, not aspirations.
 
 **Export-ready (shipped + tested in the private monorepo, 2026-07-13):**
@@ -116,15 +124,15 @@ evidence TR-SEC-010 had — commits, tests, an operating track record):**
 
 - [ ] **Disposition contract** — triage agents emit a structured disposition
       (query / think / report) as a loop-contract output field, extending
-      TR-AGT-003 (deployment proposal rev 7)
+      TR-AGT-003 (private-repo deployment proposal)
 - [ ] **Agreement-rate-gated authority promotion** — the measurable form of
       the advisory-first trust ramp: agent verdicts run advisory while
       human-agreement rate is measured; promotion to blocking/trusted cites
-      the rate, rule by rule, never the whole queue at once (rev 7)
+      the rate, rule by rule, never the whole queue at once (private-repo deployment proposal)
 - [ ] **Agent-ops metric floor** — dwell time (anomaly → human awareness),
       coverage (fraction of agent outputs a human reviewed), and
       explainability-by-trigger-ID (every agent output cites the ID of its
-      triggering event, a mandatory loop-contract field) (rev 7)
+      triggering event, a mandatory loop-contract field) (private-repo deployment proposal)
 - [ ] **Compartmentalization worked example** — the most-exposed agent gets
       the fewest permissions (write-only into a quarantine zone), the
       most-privileged agent gets no public egress, and promotion out of
@@ -143,7 +151,7 @@ evidence TR-SEC-010 had — commits, tests, an operating track record):**
 - [ ] **Layering rule** for agent/platform rollouts (operating-model doc):
       foundational ops/observability ships first; every subsequent phase is
       sized to be immediately usable — no functionality that idles behind
-      unmet dependencies (rev 7)
+      unmet dependencies (private-repo deployment proposal)
 
 ## Non-goals
 

scripts/agent-permission-guard.py

@@ -56,11 +56,20 @@
     "pip install", "npm install", "bash -c", "python -c", "python3 -c",
     "git push", "curl", "cat >", "rm -rf", "sudo",
 )
+
+
+def _verb_pattern(verb: str) -> str:
+    # Tolerate case and irregular whitespace (tabs, doubled spaces) between
+    # tokens -- a grant string shouldn't dodge FORBIDDEN by formatting alone.
+    return r"\s+".join(re.escape(token) for token in verb.split())
+
+
 FORBIDDEN_PATTERNS: list[tuple[str, re.Pattern[str]]] = [
     ("wildcard_dangerous_verb", re.compile(
-        r"(?:" + "|".join(re.escape(v) for v in _DANGEROUS_VERBS) + r").*\*"
+        r"(?:" + "|".join(_verb_pattern(v) for v in _DANGEROUS_VERBS) + r").*\*",
+        re.IGNORECASE,
     )),
-    ("unrestricted_home_read", re.compile(r"Read\(//?\*\*\)")),
+    ("unrestricted_home_read", re.compile(r"Read\(//?\*\*\)", re.IGNORECASE)),
 ]
 
 

tests/test_agent_permission_guard.py

@@ -71,6 +71,25 @@ def test_forbidden_wildcard_grant_fails_even_if_added_to_baseline(tmp_path: Path
     assert "'Bash(curl:*)'" in result.stdout
 
 
+def test_forbidden_match_is_case_and_whitespace_insensitive(tmp_path: Path) -> None:
+    settings = tmp_path / BASELINE_KEY
+    write_settings(settings, [
+        "Bash(git status:*)", "Bash(pytest:*)",
+        "Bash(PIP INSTALL:*)", "Bash(pip  install:*)",
+    ])
+    result = subprocess.run(
+        [sys.executable, str(GUARD), "--settings", BASELINE_KEY],
+        capture_output=True,
+        text=True,
+        cwd=tmp_path,
+    )
+    assert result.returncode == 1
+    forbidden_lines = [l for l in result.stdout.splitlines() if l.startswith("FORBIDDEN")]
+    assert len(forbidden_lines) == 2
+    assert "'Bash(PIP INSTALL:*)'" in result.stdout
+    assert "'Bash(pip  install:*)'" in result.stdout
+
+
 def test_unreviewed_grant_not_matching_baseline_key_fails(tmp_path: Path) -> None:
     other_path = "some/other/settings.json"
     settings = tmp_path / other_path