chore(docs): challenge the corpus with a product that has no agent record

[yordis]

• Both corpora had frozen their conclusions without ever meeting a product that omits the agent record entirely, so the stated invariants risked being artifacts of the sample rather than real convergence.
• IronClaw is that counterexample, and its evidence forces the synthesis to separate roles that must exist from resources that happen to hold them, and to name lost-update safety as distinct from append-only-ness.
• Its revisions are marked inline rather than folded silently, so a reader can still see which claims were decision-time input and which were revised after.
• Corpus counts in the prose were dropped because they force an unrelated edit every time a dossier lands.

[cursor]

PR Summary

Low Risk
Documentation-only research corpus updates; no runtime, API, or infrastructure behavior changes.

Overview
Adds IronClaw (NEAR AI) to the shared research source list and to both corpora (agent-platform product dossier plus session-store dossier), with index entries wired in.

The main substantive change is post-freeze synthesis revision, not silent rewrites: both agent-platform/synthesis.md and session-store/synthesis.md gain a front-matter note that IronClaw arrived after the synthesis was frozen, plus inline “Revised after IronClaw” blocks on convergences, divergences, comparison tables, working definitions, and design decisions. Agent-platform revisions stress that there can be no stored agent object (AgentId as scope only, persona in files, runtime on ResolvedRunProfile per run), that the definition/execution/memory trio is roles not necessarily resources, and implications for pinning, authority layers, subagents, and session vs turn-run split. Session-store revisions add a CAS-versioned record-set pole on the append-vs-mutable axis, credit IronClaw for write-boundary OCC without event sourcing, and extend cascade/retention/idempotency guidance.

Housekeeping in the same PR: corpus prose drops fixed product counts (so new dossiers do not force count churn), session-store index status text is generalized, and citation-verification wording no longer hard-codes “eighteen products.”

^{Reviewed by Cursor Bugbot for commit 5560fc3. Bugbot is set up for automated code reviews on this repo. Configure here.}

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.

[coderabbitai]

Warning

Review limit reached

@yordis, you've reached your PR review limit, so we couldn't start this review.

Next review available in: 48 minutes

You've used all free OSS reviews for now. Wait for the free limit to reset to keep reviewing this public repository.

How can I continue?

After more reviews become available, a review can be triggered using the @coderabbitai review command as a PR comment. Alternatively, push new commits to this PR.

To avoid repeated limits, reduce automatic review volume by pausing incremental auto-reviews earlier, using label-based review opt-in, excluding WIP or generated PR titles, or requesting reviews manually when the PR is ready. If your team needs uninterrupted high-volume reviews, an organization admin can enable usage-based reviews.

How do review limits work?

CodeRabbit enforces per-developer PR review limits for each organization. Most developers receive the normal plan review availability.

For paid Pro and Pro+ PR reviews, CodeRabbit uses adaptive limits for sustained high-volume activity. When a developer's recent PR review activity reaches the 95th percentile or higher among CodeRabbit users, additional reviews become available more gradually as earlier reviews age out of the rolling window.

Please refer docs for additional details.

Review details

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 3e6b8300-4c91-4b08-8a3c-1951f1a75b2c

📥 Commits

Reviewing files that changed from the base of the PR and between ad3bfbc and 5560fc3.

📒 Files selected for processing (4)

• docs/research/agent-platform/synthesis.md
• docs/research/session-store/index.md
• docs/research/session-store/products/ironclaw/index.md
• docs/research/session-store/synthesis.md

Walkthrough

The PR adds IronClaw to the research corpus, creates agent-platform and session-store dossiers, and updates both synthesis documents with IronClaw’s identity, runtime, storage, concurrency, lifecycle, and subagent findings.

Changes

IronClaw research integration

Layer / File(s)	Summary
Corpus registration and source context docs/research/README.md, docs/research/agent-platform/index.md, docs/research/session-store/index.md, docs/research/agent-platform/products/ironclaw.md	The research indexes include IronClaw. The agent dossier records source metadata and evidence precedence.
Agent model dossier docs/research/agent-platform/products/ironclaw.md	The dossier documents scope-based identity, resolved run profiles, authority ceilings, subagent behavior, lifecycle, persistence, relationships, and open questions.
Agent-platform synthesis updates docs/research/agent-platform/synthesis.md	The synthesis incorporates IronClaw’s scope identity, replay-relevant configuration capture, layered authority, separate thread and run semantics, and conservative child-run policy.
Session-store dossier docs/research/session-store/products/ironclaw/index.md	The dossier documents scoped threads, the SessionThreadService contract, typed messages, CAS and sequencing, replay, summaries, redaction, retention, subagents, and multi-host coordination.
Session-store synthesis updates docs/research/session-store/synthesis.md	The synthesis incorporates IronClaw’s record-set storage, CAS semantics, transcript operations, lineage, retention behavior, redaction, and coordination model.

Estimated code review effort: 3 (Moderate) | ~25 minutes

Possibly related PRs

• TrogonStack/trogonai#481: Adds related agent-platform research context and synthesis.
• TrogonStack/trogonai#499: Introduces the related session-store research corpus.
• TrogonStack/trogonai#485: Establishes related agent-platform corpus and synthesis documentation.

Suggested reviewers: copilot

Poem

I’m a rabbit with notes in a burrow so deep,
IronClaw’s agent and session facts I keep.
Scopes, runs, and CAS records align,
Syntheses hop into one design.
Research grows tidy, precise, and bright.

🚥 Pre-merge checks | ✅ 5

✅ Passed checks (5 passed)

Check name	Status	Explanation
Title check	✅ Passed	The title clearly identifies the documentation change: challenging the corpus with a product that has no agent record.
Description check	✅ Passed	The description directly explains IronClaw’s role as a counterexample and the resulting synthesis updates.
Docstring Coverage	✅ Passed	No functions found in the changed files to evaluate docstring coverage. Skipping docstring coverage check.
Linked Issues check	✅ Passed	Check skipped because no linked issues were found for this pull request.
Out of Scope Changes check	✅ Passed	Check skipped because no linked issues were found for this pull request.

✨ Finishing Touches

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Commit unit tests in branch yordis/san-jose

---

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

• X
• Mastodon
• Reddit
• LinkedIn

_{Comment @coderabbitai help to get the list of available commands.}

[coderabbitai]

Actionable comments posted: 4

🤖 Prompt for all review comments with AI agents

Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.

Inline comments:
In `@docs/research/session-store/index.md`:
- Around line 60-64: The sweep-status prose in
docs/research/session-store/index.md:60-64 must remove the hard-coded citation
total and manually maintained remaining-dossier list, replacing them with
invariant wording or repository-generated values. Also update
docs/research/agent-platform/synthesis.md:435-437 to replace “the eighteenth
product” with product-neutral wording; preserve the surrounding meaning without
embedding corpus-size-dependent counts.

In `@docs/research/session-store/products/ironclaw/index.md`:
- Around line 301-302: Update the method count in the SessionThreadService
description from 32 to 34, while keeping the stated count of 16 required methods
unchanged.

In `@docs/research/session-store/synthesis.md`:
- Line 4: Update the later sentence near the “nine dossiers” reference to remove
the fixed corpus count while preserving its meaning and surrounding wording.
- Around line 519-534: The comparison in the “Industry’s answer” paragraph
overstates what a single expected-version append provides. Narrow the claim to
lost-update and cross-record invariant protection, or explicitly state that
lease ownership, heartbeat renewal, expiry, and pre-side-effect admission
require modeled aggregate lease transitions with a lease protocol; do not imply
expected-version append alone supplies those active-run protections.

🪄 Autofix

Fix all unresolved CodeRabbit comments on this PR:

• Push a commit to this branch (recommended)
• Create a new PR with the fixes

---

ℹ️ Review info

⚙️ Run configuration

Configuration used: Organization UI

Review profile: CHILL

Plan: Pro Plus

Run ID: 5df3eb09-3636-40bc-aa6c-fdb632768d96

📥 Commits

Reviewing files that changed from the base of the PR and between 662864b and ad3bfbc.

📒 Files selected for processing (7)

• docs/research/README.md
• docs/research/agent-platform/index.md
• docs/research/agent-platform/products/ironclaw.md
• docs/research/agent-platform/synthesis.md
• docs/research/session-store/index.md
• docs/research/session-store/products/ironclaw/index.md
• docs/research/session-store/synthesis.md

[Copilot]

Copilot was unable to review this pull request because the user who requested the review has reached their quota limit.