@hassangameryt/freeflow 0.7.3 → 0.8.0

package/skills/review-artifact/SKILL.md

@@ -9,47 +9,55 @@ Judge whether the complete artifact faithfully represents accepted intent and is
 
 [Verify Work](../../skills/verify-work/SKILL.md) establishes factual support where claims need verification. Review judges fitness without creating evidence, settling owner decisions, or authorizing revisions.
 
+The **Rules** below are binding; if one cannot be met, stop and say why. Everything else is **judgment**: follow its stated goal, and adapt when a different choice serves that goal better.
+
+## Rules
+
+- Writing or revising a Spec or Plan does not require independent review. Select it only when requested, or justified and authorized, to protect a concrete boundary; loading this skill or switching compute profiles creates no independence.
+- An independent reviewer reports and stops: no edits, owner decisions, selection among materially different revisions, adjudication, or further dispatch, and no continuing until Pass. It runs a missing active check only when review authority covers that exact check; otherwise it reports Needs evidence.
+- Findings do not authorize revisions. Derive the judgment from your dispositions rather than adopting the reviewer's conclusion.
+- Never settle a user decision or revise upstream intent, accepted content, or owner decisions merely to make an artifact consistent or to obtain Pass. Never review until Pass.
+- Never request Review 4. A different reviewer, renamed use, or local revision does not reset the review budget.
+- Never present artifact fitness as acceptance, implementation, or execution authority.
+
 ## Choose The Role And Intended Use
 
+Goal: review the whole artifact, for the use it must actually serve.
+
 - **Self-review:** the author checks its own artifact before accepting or using it. This is the normal required route.
 - **Independent review:** a separately selected reviewer judges a state it did not produce, from a separate context.
 - **Adjudication:** the receiving agent evaluates a completed independent report.
 
-Writing or revising a Spec or Plan does not automatically require independent review. Select that route only when requested or justified and authorized to protect a concrete boundary. Loading this skill or switching compute profiles creates no independence.
-
-An independent reviewer reports and stops; it cannot adjudicate its own findings or become the revising agent within that role.
-
-Identify the artifact's exact state, type, intended use, accepted outcome and amendments, non-goals, governing format, source basis, unresolved questions, dependencies, and evidence gaps. Inspect the complete artifact and relevant sources directly—not only its author's narrative.
+Identify the artifact's exact state, type, intended use, accepted outcome and amendments, non-goals, governing format, source basis, unresolved questions, dependencies, and evidence gaps. Inspect the complete artifact and relevant sources directly, not only its author's narrative.
 
 ## Establish Which Sources Govern
 
-Check current intent and accepted upstream requirements before dependent artifacts. Follow material amendments, supersession, and deliberately selected experimental findings. Do not infer acceptance from a proposal, a review label, a passing test, or an unexplained "user approved" summary.
-
-Distinguish accepted behavior from the suggested implementation mechanism. Check both directions:
+Goal: judge against current accepted intent, not the artifact's own baseline.
 
-- Has the artifact introduced stronger obligations than the accepted outcome needs?
-- Has it omitted accepted changes or retained a superseded implementation baseline?
+Check current intent and accepted upstream requirements before dependent artifacts. Follow material amendments, supersession, and deliberately selected experimental findings. Do not infer acceptance from a proposal, a review label, a passing test, or an unexplained "user approved" summary.
 
-An experiment can establish a fact without selecting it for production. Conversely, an implementation Plan cannot ignore later accepted learning merely because its older version was reviewed.
+Distinguish accepted behavior from the suggested implementation mechanism, and check both directions: has the artifact introduced stronger obligations than the accepted outcome needs, and has it omitted accepted changes or kept a superseded baseline? An experiment can establish a fact without selecting it for production; an implementation Plan cannot ignore later accepted learning because its older version was reviewed.
 
-If upstream content is unresolved or contradictory, mark dependent material contingent. Do not generate exhaustive downstream findings against an unsettled premise or rewrite the upstream requirement to match the artifact.
+If upstream content is unresolved or contradictory, mark dependent material contingent rather than generating exhaustive downstream findings against an unsettled premise.
 
 Read [Track Work](../../skills/track-work/SKILL.md) before reviewing a Working Record. Use [Action Selection](../../skills/action-selection/SKILL.md) when source inspection could branch broadly; skip it for an obvious artifact or dependency read.
 
 ## Judge The Artifact By Its Job
 
-- **Working Record:** accurate current task memory, recoverable decisions and work, evidence limits, provisional remaining route, and one next useful action.
+Goal: ask of each artifact only what its job requires.
+
+- **Working Record:** accurate current task memory, recoverable decisions and work, evidence limits, provisional remaining route, defining artifacts and Recovery sources sufficient for a fresh context to resume, and one next useful action.
 - **Spec or content contract:** the accepted behavior, boundaries, evidence, and uncertainty needed for its intended use.
 - **Plan:** a supported ordered strategy, dependencies, assumptions, checks, and invalidation conditions.
 - **Decision record or ADR:** the choice, owner, source, alternatives, rationale, consequences, and revisit or supersession conditions.
 - **Handoff:** enough truthful point-in-time context for safe continuation without replacing the live record.
 - **Other artifact:** its declared purpose without taking over another artifact's job.
 
-A proposal can be fit for discussion without being implementation-ready. Do not block on an open question that the artifact intentionally preserves and whose answer is unnecessary for its current use.
+A proposal can be fit for discussion without being implementation-ready. Do not block on an open question the artifact intentionally preserves and whose answer its current use does not need.
 
 ## Apply The Shared Kernel
 
-Use only lenses capable of changing fitness:
+Goal: use only the lenses that can change fitness.
 
 - **Source alignment:** accepted requirements, amendments, decisions, and live facts agree with the artifact.
 - **Sufficiency:** it contains enough for its use without implying every future question is settled.
@@ -58,12 +66,14 @@ Use only lenses capable of changing fitness:
 - **Behavior and failure:** consequential states, forbidden outcomes, observers, and recovery are defined where required.
 - **Dependency integrity:** relevant upstream and downstream sources agree, or their contingency is explicit.
 - **Scope and minimality:** requirements and process are justified rather than speculative or mechanically exhaustive.
-- **Clarity and continuity:** a future reader can use the artifact without reconstructing the author's private reasoning or full transcript.
+- **Clarity and continuity:** a future reader can use the artifact without reconstructing the author's private reasoning or the full transcript.
 
 Do not demand implementation detail from a PRD, progress history from a Plan, or universal guarantees from a bounded prototype report. Finding no material issue is valid.
 
 ## Classify Material Findings
 
+Goal: every reported item is supported and names the right kind of problem.
+
 Ask in order:
 
 1. Does intended-use fitness depend on unsettled content or an owner choice? Report a **Question**.
@@ -71,72 +81,63 @@ Ask in order:
 3. Is there a supported defect requiring revision for this use? Report a **Blocking Issue** if the boundary cannot safely be crossed, or a **Non-blocking Issue** if revision can safely be deferred.
 4. Would it only improve a different or broader use? Omit it by default, or identify an **Improvement** when materially relevant or requested.
 
-An Issue identifies the exact location or dependency, violated source or artifact responsibility, evidence, and concrete consequence. Explain why a Blocking Issue prevents the intended use. Recommend a specific revision only when source intent and dependency consequences support it.
+An Issue identifies the exact location or dependency, violated source or artifact responsibility, evidence, and concrete consequence; explain why a Blocking Issue prevents the intended use. Recommend a specific revision only when source intent and dependency consequences support it.
 
-Needs evidence identifies the load-bearing claim, required observer, existing evidence and limits, why the gap matters, and the smallest observation that could disagree. Establish that the claim is required before recommending more proof; optional stronger claims may be qualified rather than expanding the task. Distinguish missing evidence from a demonstrated artifact defect. Check which source/check state a reported pass examined and whether later changes reduced acceptance or invalidated its applicability; a newer report cannot repair that mismatch.
+Needs evidence identifies the load-bearing claim, required observer, existing evidence and limits, why the gap matters, and the smallest observation that could disagree. Establish that the claim is required before recommending more proof; optional stronger claims may be qualified rather than expanding the task. Distinguish missing evidence from a demonstrated defect, and check which source or check state a reported pass examined and whether later changes reduced acceptance or invalidated it; a newer report cannot repair that mismatch.
 
 Do not classify wording preference, exhaustive edge cases, intentional deferrals, or polished presentation as defects.
 
 ## Self-Review Once
 
-Apply the kernel silently to the artifact you produced after its load-bearing facts have adequate support. Do not create formal review items, a numbered review, or an independent judgment for this route.
+Goal: make the artifact fit for its use without turning review into open-ended rewriting.
 
-For a clear covered defect whose revision basis is settled:
+Apply the kernel silently to the artifact you produced once its load-bearing facts have adequate support; do not create formal review items, a numbered review, or an independent judgment for this route.
 
-1. return to the authoring activity and revise it;
-2. reconcile affected dependencies within authority;
-3. re-verify affected factual claims;
-4. repeat source alignment and affected review lenses once.
+For a clear covered defect whose revision basis is settled: return to the authoring activity and revise it, reconcile affected dependencies within authority, re-verify affected factual claims, and repeat source alignment and the affected lenses once.
 
-Do not settle a user decision or revise upstream intent merely to make the artifact consistent. Use [Discuss](../../skills/discuss/SKILL.md) for changed direction or strategy, [Decision Gate](../../skills/decision-gate/SKILL.md) for a blocking owner choice or source conflict, and [Diagnose Failure](../../skills/diagnose-failure/SKILL.md) for unsupported or repeated causal claims. Return material coordination changes through [Workflow](../../skills/workflow/SKILL.md).
+Use [Discuss](../../skills/discuss/SKILL.md) for changed direction or strategy, [Decision Gate](../../skills/decision-gate/SKILL.md) for a blocking owner choice or source conflict, and [Diagnose Failure](../../skills/diagnose-failure/SKILL.md) for unsupported or repeated causal claims. Return material coordination changes through [Workflow](../../skills/workflow/SKILL.md).
 
-Stop when the artifact is fit for its agreed use or a material unresolved issue is routed. Self-review is not a reason to add independent review automatically.
+Stop when the artifact is fit for its agreed use or a material unresolved issue is routed. Self-review is not a reason to add independent review.
 
 ## Perform A Selected Independent Review
 
-Before preparing or performing it, read [Independent Artifact Reviewer Contract](references/independent-artifact-reviewer-contract.md).
-
-Use available evidence. Run a missing active check only when review authority covers that exact check; otherwise report Needs evidence. Apply Verify Work to covered checks while retaining ownership of the fitness judgment.
+Goal: a separate context judges the artifact from the evidence and reports once.
 
-Inspect and report without edits, owner decisions, selection among materially different revisions, adjudication, or another review dispatch. The independent review ends with its report:
+Before preparing or performing it, read [Independent Artifact Reviewer Contract](references/independent-artifact-reviewer-contract.md). Use available evidence and apply Verify Work to covered checks while keeping ownership of the fitness judgment. The review ends with its report:
 
 - **Blocking:** at least one Blocking Issue.
 - **Inconclusive:** no Blocking Issue, but a material Question or Needs evidence prevents judgment.
 - **Non-blocking:** only Non-blocking Issues remain.
 - **Pass:** no Issues or material unresolved items remain.
 
-Improvements do not change the judgment or authorize revision. Do not continue merely to obtain Pass.
+Improvements do not change the judgment or authorize revision.
 
 ## Adjudicate Without Inheriting The Verdict
 
-Read [Adjudicate Artifact Review](references/adjudicate-artifact-review.md) before adjudicating.
+Goal: decide what the report actually establishes before acting on it.
 
-Assess each material item against the actual artifact, source truth, dependencies, and evidence:
+Read [Adjudicate Artifact Review](references/adjudicate-artifact-review.md) before adjudicating. Assess each material item against the actual artifact, source truth, dependencies, and evidence:
 
 - **Accepted:** supported and applicable.
 - **Rejected:** unsupported, stale, resolved, duplicate, preference-only, outside the artifact's job, or a source misread.
 - **Open:** evidence or an owner decision is missing.
 
-Derive the overall judgment from those dispositions rather than separately accepting the reviewer's conclusion. Findings do not authorize revisions.
-
 Pass permits the intended use. Non-blocking permits it with explicit deferrals. Inconclusive requires the missing evidence or decision. Blocking prevents the disputed use and returns the narrowest supported revision or reconsideration to its owner through Workflow.
 
 When revision is ready, state the supported problem, revision basis, rationale, dependency impact, observing boundary, authority, and whether focused independent follow-up remains justified.
 
 ## Preserve The Independent Review Budget
 
+Goal: independent judgment is spent where it changes the outcome.
+
 For one reviewed state and intended-use boundary:
 
 - **Review 1:** first selected independent review, broad by default.
 - **Review 2:** separately selected focused follow-up when accepted revisions, affected dependencies, new evidence, or remaining risk warrant it.
 - **Review 3:** exceptional, separately authorized, and final, after the revision basis and dependency boundary are understood.
 
-A revision does not automatically authorize another review. If Review 2 remains Blocking, diagnose repeated or unclear causes; return an independent clear local artifact defect to its owner when diagnosis adds no value. Do not automatically revise and dispatch Review 3.
-
-Do not request Review 4. A different reviewer, renamed use, or local revision does not reset the budget. Workflow may establish a new cycle only for a materially new artifact state and intended-use boundary. Self-review does not consume the budget.
+A revision does not authorize another review. If Review 2 remains Blocking, diagnose repeated or unclear causes; return an independent clear local defect to its owner when diagnosis adds no value, rather than revise and dispatch Review 3 automatically. Workflow may start a new cycle only for a materially new artifact state and intended-use boundary. Self-review does not consume the budget.
 
 ## Stop
 
-Self-review ends with supported fitness or a routed issue. Independent review ends with its report. Adjudication ends with material dispositions and an explicit next route.
-
-Do not review until Pass, turn optional improvement into unfinished work, or present artifact fitness as acceptance, implementation, or execution authority.
+Self-review ends with supported fitness or a routed issue. Independent review ends with its report. Adjudication ends with material dispositions and an explicit next route. Do not turn optional improvement into unfinished work.

package/skills/review-artifact/references/adjudicate-artifact-review.md

@@ -2,7 +2,7 @@
 
 Read this after an independent artifact review returns and before selecting what follows from its report.
 
-Adjudication belongs to the receiving active agent. It tests the report against the artifact, source truth, dependencies, and evidence. It is not independent review and does not inherit the reviewer's authority or judgment.
+This reference is judgment; Review Artifact's Rules still bind. Adjudication belongs to the receiving active agent. It tests the report against the artifact, source truth, dependencies, and evidence. It is not independent review and does not inherit the reviewer's authority or judgment.
 
 ## Dispose Each Material Item
 

package/skills/review-artifact/references/independent-artifact-reviewer-contract.md

@@ -2,7 +2,7 @@
 
 Read this before preparing or performing a separately selected independent review of a durable artifact.
 
-The reviewer did not produce the reviewed artifact state. Give them the complete artifact, source truth, dependencies, and evidence needed to judge it directly. Do not provide only the producer's summary, reasoning, or change description.
+This reference is a contract: the required context and the reviewer prompt are binding as written. The reviewer did not produce the reviewed artifact state. Give them the complete artifact, source truth, dependencies, and evidence needed to judge it directly. Do not provide only the producer's summary, reasoning, or change description.
 
 ## Required Context
 

package/skills/review-work/SKILL.md

@@ -9,29 +9,39 @@ Judge whether the actual work satisfies the accepted outcome, remains proportion
 
 [Verify Work](../../skills/verify-work/SKILL.md) establishes what observations prove. Review judges their sufficiency and the resulting work; it does not replace verification, choose user-owned behavior, or authorize corrections.
 
+The **Rules** below are binding; if one cannot be met, stop and say why. Everything else is **judgment**: follow its stated goal, and adapt when a different choice serves that goal better.
+
+## Rules
+
+- Independent review is never automatic after implementation or self-review. Select it through [Workflow](../../skills/workflow/SKILL.md) only when requested, or justified and authorized, to protect a concrete risk, dependency, integration, or delivery boundary. A review budget permits no dispatch by itself, and switching models or loading this skill does not create independence.
+- An independent reviewer reports and stops: no edits, no choosing among materially different remedies, no adjudication, no further dispatch, and no continuing until Pass. It runs a missing active check only when review authority covers that exact check; otherwise it reports Needs evidence.
+- Findings are evidence, not commands or correction authority. Derive the judgment from your dispositions; never adopt the reviewer's verdict wholesale. Actual correction belongs to [Execute Work](../../skills/execute-work/SKILL.md) under covered authority.
+- Never lower acceptance or change tests, Specs, policies, or source truth to obtain Pass, and never review until Pass.
+- Never request Review 4. Renaming scope, changing reviewers, or making a local correction does not reset the review budget.
+
 ## Choose The Review Role
 
+Goal: the review done is the one the boundary needs.
+
 - **Self-review:** the producer checks its own supported result before accepting or reusing it. This is the normal required route.
 - **Independent review:** a separately selected reviewer judges a state it did not produce, from a separate context.
 - **Adjudication:** the receiving agent checks a returned independent report and decides what its findings establish.
 
-An independent reviewer reports and stops. It does not adjudicate its own findings or become the fixing agent. Switching models or loading this skill does not create independence.
-
-Independent review is not automatic after implementation or self-review. Select it through Workflow only when requested or justified and authorized to protect a concrete risk, dependency, integration, or delivery boundary. A review budget permits no dispatch by itself.
-
 ## Establish What The Review Protects
 
+Goal: judge the actual state against its real sources, not the producer's account of it.
+
 Identify the exact code, diff, commit, or integrated state; its accepted outcome, non-goals, relevant requirements and amendments; the future action being protected; and the verification evidence and gaps.
 
-Inspect the actual state and governing sources—not just the producer's summary or explanation. A detailed execution contract can itself have been mistaken. Check the user outcome before judging compliance with the suggested implementation steps.
+Inspect the actual state and governing sources, not just the producer's summary or explanation. A detailed execution contract can itself have been mistaken; check the user outcome before judging compliance with the suggested implementation steps.
 
 Before reviewing changes to trust boundaries, authentication, authorization, permissions, untrusted input, secrets, sensitive data, security-relevant dependencies, code execution, or security-sensitive failure behavior, read [Security Risk Lens](references/security-risk-lens.md).
 
-Use [Action Selection](../../skills/action-selection/SKILL.md) for uncertain or broad inspection. An obvious diff or source read does not need another selection exercise.
+Use [Action Selection](../../skills/action-selection/SKILL.md) for uncertain or broad inspection; an obvious diff or source read needs no selection exercise.
 
 ## Apply The Same Technical Kernel
 
-Use lenses that can change the boundary judgment:
+Goal: use only the lenses that can change the boundary judgment.
 
 - **Alignment:** required behavior and accepted amendments are implemented without omission or invented obligations.
 - **Correctness and integration:** reachable paths, affected callers, and components work together without regressions.
@@ -40,12 +50,12 @@ Use lenses that can change the boundary judgment:
 - **Design and minimality:** complexity and coordination serve requirements or demonstrated needs rather than optional stronger guarantees.
 - **Maintainability:** behavior and failure paths remain understandable and changeable.
 
-Do not assume more implementation is better. Check both missing accepted work and unnecessary additions. A prototype result is not production acceptance, and a test protecting a mechanism does not make that mechanism required.
-
-If the implementation follows an artifact whose baseline conflicts with a later accepted change, expose that source mismatch. Do not choose whichever source makes the work pass or treat every unselected experiment as a production obligation.
+More implementation is not better: check both missing accepted work and unnecessary additions. A prototype result is not production acceptance, and a test protecting a mechanism does not make that mechanism required. If the implementation follows an artifact whose baseline conflicts with a later accepted change, expose that source mismatch rather than choose whichever source makes the work pass.
 
 ## Classify Before Calling Something A Defect
 
+Goal: every reported item is supported and names the right kind of problem.
+
 Ask in order:
 
 1. Is the relevant behavior or source truth settled? If not, report a **Question** when it affects the boundary.
@@ -53,29 +63,29 @@ Ask in order:
 3. Does this boundary require correction? If it cannot safely be crossed, report a **Blocking Issue**. If correction is required but safely deferrable, report a **Non-blocking Issue**.
 4. Is it merely useful beyond this boundary? Omit it by default, or identify an **Improvement** when materially relevant or explicitly requested.
 
-An Issue needs an exact location or path, violated requirement or invariant, supporting evidence, and concrete consequence. A Blocking Issue must explain why the protected boundary cannot be crossed.
+An Issue needs an exact location or path, the violated requirement or invariant, supporting evidence, and a concrete consequence; a Blocking Issue must explain why the protected boundary cannot be crossed.
 
-Needs evidence must identify the claim, required observer, existing evidence and limit, why the gap matters, and the smallest observation that could disagree. First establish that the stronger claim is required; missing proof of an optional guarantee does not automatically justify more work. Check whether supplied observations apply to the reviewed candidate and whether material assertion changes preserved the required property. A missing or stale result is not itself a code defect; recover evidence or establish the observer before prescribing a fix.
+Needs evidence identifies the claim, required observer, existing evidence and its limit, why the gap matters, and the smallest observation that could disagree. First establish that the stronger claim is required; missing proof of an optional guarantee does not justify more work. Check whether supplied observations apply to the reviewed candidate and whether material assertion changes preserved the required property. A missing or stale result is not itself a code defect: recover evidence or establish the observer before prescribing a fix.
 
 Do not turn preferences, imagined edge cases, intentional deferrals, or hypothetical completeness into Issues. Finding no material issue is valid.
 
 ## Self-Review And Re-enter Narrowly
 
-Apply the kernel silently after evidence initially supports the produced result. Do not create formal review items, a numbered review cycle, or an independent-review judgment for self-review.
+Goal: catch real defects in your own result without turning review into an open-ended editing loop.
+
+Apply the kernel silently after evidence initially supports the produced result; do not create formal review items, a numbered review cycle, or an independent-review judgment for self-review.
 
-For one clear covered defect, return to the producer, correct it, re-verify the affected boundary, and repeat only affected review lenses once. Do not restart every check or add a reviewer because a local correction occurred.
+For one clear covered defect, return to the producer, correct it, re-verify the affected boundary, and repeat only the affected lenses once. Do not restart every check or add a reviewer because a local correction occurred.
 
-When the outcome remains settled but the approach is invalidated, return for implementation preparation. Use [Diagnose Failure](../../skills/diagnose-failure/SKILL.md) for unclear or repeatedly failing causes, and [Decision Gate](../../skills/decision-gate/SKILL.md) for unsettled accepted behavior or a user-owned choice. Return coordination or scope changes to [Workflow](../../skills/workflow/SKILL.md).
+When the outcome remains settled but the approach is invalidated, return for implementation preparation. Use [Diagnose Failure](../../skills/diagnose-failure/SKILL.md) for unclear or repeatedly failing causes, [Decision Gate](../../skills/decision-gate/SKILL.md) for unsettled accepted behavior or a user-owned choice, and Workflow for coordination or scope changes.
 
 Stop self-review when the supported state is accepted or a material issue has a clear next route. Optional polish is not permission to continue editing.
 
 ## Perform A Selected Independent Review
 
-Before preparing or performing it, read [Independent Work Reviewer Contract](references/independent-work-reviewer-contract.md).
+Goal: a separate context judges the state from the evidence and reports once.
 
-Use supplied and available evidence. Do not run a missing active check unless review authority covers that exact check; otherwise report Needs evidence. A covered active check uses Verify Work's check-result and claim-result semantics without changing review ownership.
-
-Inspect and report without edits, choosing among materially different remedies, adjudication, another review dispatch, or continuing until Pass. Recommend a specific correction only when supported rather than merely plausible.
+Before preparing or performing it, read [Independent Work Reviewer Contract](references/independent-work-reviewer-contract.md). Use supplied and available evidence; a covered active check uses Verify Work's check-result and claim-result semantics without changing review ownership. Recommend a specific correction only when supported rather than merely plausible.
 
 The independent judgment is:
 
@@ -88,34 +98,30 @@ Improvements do not change the judgment or authorize implementation. Every judgm
 
 ## Adjudicate The Returned Report
 
-Read [Adjudicate Work Review](references/adjudicate-work-review.md) before adjudicating.
+Goal: decide what the report actually establishes before acting on it.
 
-Check each material finding against the actual reviewed state, sources, and evidence:
+Read [Adjudicate Work Review](references/adjudicate-work-review.md) before adjudicating. Check each material finding against the actual reviewed state, sources, and evidence:
 
 - **Accepted:** supported and applicable.
 - **Rejected:** unsupported, stale, already resolved, duplicate, preference-only, out of scope, or based on a source misread.
 - **Open:** evidence or a decision is missing.
 
-Derive the overall judgment from those dispositions; do not separately accept the reviewer's verdict. Findings are evidence, not commands or correction authority.
-
 Pass permits proceeding at the reviewed boundary. Non-blocking permits proceeding with explicit deferrals. Inconclusive requires the missing evidence or decision. Blocking prevents crossing the boundary and returns the smallest supported correction or reconsideration to its owner.
 
-When correction is ready, state the supported problem, remedy, rationale, verification boundary, authority, and whether independent follow-up is still warranted. Actual correction belongs to [Execute Work](../../skills/execute-work/SKILL.md), not the independent reviewer.
+When correction is ready, state the supported problem, remedy, rationale, verification boundary, authority, and whether independent follow-up is still warranted.
 
 ## Bound Independent Review
 
-Keep the existing budget for the same state and boundary:
+Goal: independent judgment is spent where it changes the outcome.
+
+For the same state and boundary:
 
 - **Review 1:** first selected independent review, broad by default.
 - **Review 2:** separately selected focused follow-up when corrections, affected interactions, new evidence, or remaining risk still require independent judgment.
-- **Review 3:** exceptional, separately authorized, and final; use only after the cause and correction boundary are understood.
-
-A fix does not automatically require another review. If Review 2 remains Blocking, diagnose a repeated, extending, or unclear cause; return an unrelated clear local defect to its owner when diagnosis would add no information. Do not automatically fix and dispatch Review 3.
+- **Review 3:** exceptional, separately authorized, and final; only after the cause and correction boundary are understood.
 
-Do not request Review 4. Renaming scope, changing reviewers, or making a local correction does not reset the budget. Workflow may establish a new cycle only for a materially new reviewed state and boundary. Self-review does not consume this budget.
+A fix does not automatically require another review. If Review 2 remains Blocking, diagnose a repeated, extending, or unclear cause; return an unrelated clear local defect to its owner when diagnosis would add no information, rather than fix and dispatch Review 3 automatically. Workflow may start a new cycle only for a materially new reviewed state and boundary. Self-review does not consume this budget.
 
 ## Stop At The Selected Result
 
-Self-review ends with acceptance or a routed issue. Independent review ends with its report. Adjudication ends with dispositions and an explicit next route.
-
-Do not review until Pass, lower acceptance to obtain it, or turn optional improvement into unfinished work. Preserve the agreed user-facing return boundary and any unresolved limits.
+Self-review ends with acceptance or a routed issue. Independent review ends with its report. Adjudication ends with dispositions and an explicit next route. Do not turn optional improvement into unfinished work; preserve the agreed user-facing return boundary and any unresolved limits.

package/skills/review-work/references/adjudicate-work-review.md

@@ -2,7 +2,7 @@
 
 Read this after an independent work review returns and before selecting what follows from its report.
 
-Adjudication belongs to the receiving active agent. It tests the report against the reviewed state, source truth, and evidence. It is not independent review and does not inherit the reviewer's authority or judgment.
+This reference is judgment; Review Work's Rules still bind. Adjudication belongs to the receiving active agent. It tests the report against the reviewed state, source truth, and evidence. It is not independent review and does not inherit the reviewer's authority or judgment.
 
 ## Dispose Each Material Item
 

package/skills/review-work/references/independent-work-reviewer-contract.md

@@ -2,7 +2,7 @@
 
 Read this before preparing or performing a separately selected independent review of implementation or integrated work.
 
-The reviewer did not produce the reviewed state. Give them the work product, source truth, and evidence needed to judge it directly. Do not provide only the producer's summary, reasoning, or claimed result.
+This reference is a contract: the required context and the reviewer prompt are binding as written. The reviewer did not produce the reviewed state. Give them the work product, source truth, and evidence needed to judge it directly. Do not provide only the producer's summary, reasoning, or claimed result.
 
 ## Required Context
 

package/skills/review-work/references/security-risk-lens.md

@@ -2,7 +2,7 @@
 
 Read this before reviewing work that changes trust boundaries, authentication, authorization, permissions, untrusted input, secrets, sensitive data, security-relevant dependencies or external integrations, code-execution boundaries, or failure behavior with security consequences.
 
-This lens helps frame review. It does not replace repository security policy, threat modeling, specialist review, or stack-specific hardening guidance.
+This reference is judgment, with two Rules: security-sensitive product behavior stays user-owned, so never invent access, retention, logging, or fail-open or fail-closed policy during review; and never copy secrets, credentials, unrestricted personal data, or sensitive payloads into review context. It helps frame review and does not replace repository security policy, threat modeling, specialist review, or stack-specific hardening guidance.
 
 ## Establish Authority
 
@@ -14,8 +14,6 @@ Identify:
 - accepted security and privacy policy and threat model;
 - compatibility and failure behavior that security controls must preserve.
 
-Security-sensitive product behavior remains user-owned. Do not invent access, retention, logging, or fail-open or fail-closed policy during review.
-
 ## Trace The Paths
 
 Review the affected path from boundary to effect:
@@ -45,7 +43,7 @@ Prefer evidence at the real boundary:
 
 A generic scanner pass cannot prove authorization logic or policy correctness. A unit test cannot prove deployed headers, identity configuration, or infrastructure permissions.
 
-Do not copy secrets, credentials, unrestricted personal data, or sensitive payloads into review context. Use the smallest sanitized evidence that supports the item.
+Use the smallest sanitized evidence that supports the item.
 
 ## Calibrate Items
 

package/skills/setup-freeflow/SKILL.md

@@ -54,7 +54,9 @@ Do not create or append Freeflow instructions in `AGENTS.md`, `CLAUDE.md`, `.cla
 
 ## Add Personal Overrides Only When Requested
 
-Use `.freeflow/local.json` for an explicitly selected per-checkout override of `enabled`, `contextVirtualization`, or `conversationHistory`. Omitted values inherit from repository config and built-in defaults. Configure optional Cognitive Routing values only through its explicit settings surface.
+Use `.freeflow/local.json` for an explicitly selected per-checkout override of `enabled`. Omitted values inherit from repository config and built-in defaults. Configure optional Cognitive Routing values only through its explicit settings surface.
+
+The legacy context keys `contextVirtualization` and `conversationHistory` are no longer supported. Delete either key from `.freeflow/config.json` or `.freeflow/local.json` wherever present; the containing file is invalid until repaired. Setup never removes them automatically.
 
 In Pi, `/freeflow settings` edits personal overrides and `/freeflow settings repo` edits shared settings. Prefer those controls when available.
 

package/skills/setup-freeflow/references/activation-contract.md

@@ -10,13 +10,9 @@ Freeflow uses layered checkout configuration. Shared repository activation, pers
 {}
 ```
 
-Missing core keys use built-in defaults:
+Missing core keys use the built-in default `enabled: true`.
 
-- `enabled: true`
-- `contextVirtualization: false`
-- `conversationHistory: false`
-
-Optional capability configuration is explicit and remains independently gated. The legacy `defaultMode`, `interactionContract`, and `skills` keys are unsupported and make the configuration invalid.
+Optional Cognitive Routing and Tool Execution configuration is explicit and remains independently gated. The legacy `defaultMode`, `interactionContract`, and `skills` keys are unsupported and make the configuration invalid. The removed `contextVirtualization` and `conversationHistory` keys are also unsupported in either config layer; delete them manually wherever present before Freeflow can activate. Setup does not migrate or remove them automatically.
 
 Do not add task state, phase, file inventories, plans, version metadata, activation paths, empty optional sections, host instruction copies, or generated workflow text.
 
@@ -26,17 +22,15 @@ Missing or invalid repository config means repository activation is absent. A va
 
 `.freeflow/local.json` is optional per-checkout state. It cannot activate Freeflow without a valid repository config.
 
-It may override these core values:
+It may override the core enablement value:
 
 ```json
 {
-  "enabled": false,
-  "contextVirtualization": true,
-  "conversationHistory": true
+  "enabled": false
 }
 ```
 
-Every field is optional. Omission inherits the repository value, then the built-in default. Local core precedence is property-level, not whole-file replacement.
+The field is optional. Omission inherits the repository value, then the built-in default. Local core precedence is property-level, not whole-file replacement.
 
 Keep local config untracked and ignored. In a Git checkout, prefer `.git/info/exclude` or an existing ignore rule over an unsolicited shared `.gitignore` edit. Refuse to write personal overrides to a tracked local file.
 
@@ -50,7 +44,7 @@ Resolve remaining configurable values in this order:
 host session override -> personal override -> repository value -> built-in default
 ```
 
-Host session overrides are host state, not config, and cannot bypass missing or invalid repository activation. Pi may temporarily override Freeflow enablement and the remaining optional context capabilities. Cognitive Routing has its own host-managed session control.
+Host session overrides are host state, not config, and cannot bypass missing or invalid repository activation. Pi may temporarily override Freeflow enablement. Cognitive Routing has its own host-managed session control.
 
 Effective Freeflow requires:
 
@@ -59,7 +53,7 @@ Effective Freeflow requires:
 3. session-resolved `enabled: true`;
 4. the mandatory `core.md` and `interaction-contract.md` prompt files.
 
-The core prompt and Interaction Contract are always delivered together when Freeflow is enabled. Base Freeflow skills are exposed with that core surface. Context Virtualization, Conversation History, and Cognitive Routing remain independently gated capabilities.
+The core prompt and Interaction Contract are always delivered together when Freeflow is enabled. Base Freeflow skills are exposed with that core surface. Cognitive Routing and Tool Execution remain independently gated capabilities; Context Virtualization and Conversation History are removed.
 
 ## Mutation Authority By State
 

package/skills/setup-freeflow/references/host-setup.md

@@ -28,15 +28,15 @@ Pi provides these settings scopes:
 
 ```text
 /freeflow settings          personal overrides
-/freeflow settings session  temporary enablement and optional-context overrides
+/freeflow settings session  temporary Freeflow and Cognitive Routing session controls
 /freeflow settings repo     shared repository settings
 ```
 
-Session settings can temporarily inherit, enable, or disable Freeflow, Context Virtualization, and Conversation History. One reset action clears those remaining session overrides. They persist in the active Pi session branch JSONL, never write `.freeflow/config.json` or `.freeflow/local.json`, and cannot bypass missing or invalid repository activation. Master changes reload Pi resources when supported.
+Session settings can temporarily inherit, enable, or disable Freeflow and override Cognitive Routing profiles/delegation. One reset action clears those session overrides. They persist in the active Pi session branch JSONL, never write `.freeflow/config.json` or `.freeflow/local.json`, and cannot bypass missing or invalid repository activation. Master changes reload Pi resources when supported.
 
 Personal settings use omission to inherit. The settings path refuses to overwrite invalid or tracked local config and establishes local Git exclusion when needed. These are user-operated host controls; they update state before the next model turn. Agent-performed file edits still follow Setup Freeflow's mutation-authority rules.
 
-When effective, Pi's `before_agent_start` path preserves the existing system prompt and composes the mandatory `core.md` and `interaction-contract.md` fragments plus effective optional capability prompts. Runtime State is refreshed at session start, after context reconstruction or loss, and when displayed state changes; unchanged state remains in the provider context. Base skill bodies and capability bodies remain discoverable rather than hidden persistent messages. Cognitive Routing, Context Virtualization, and Conversation History are independently gated.
+When effective, Pi's `before_agent_start` path preserves the existing system prompt and composes the mandatory `core.md` and `interaction-contract.md` fragments plus the Cognitive Routing prompt when effective. Runtime State is refreshed at session start, after context reconstruction or loss, and when displayed state changes; unchanged state remains in the provider context. Base skill bodies and capability bodies remain discoverable rather than hidden persistent messages. The former Context Virtualization and Conversation History settings, prompts, and tool operations are removed.
 
 A config written during setup is visible to the available Pi adapter on the next agent turn. After first activation, tell the user to run `/reload` so Pi refreshes skills and resources fully. Without reload, the adapter can load core runtime context on the next agent turn, but resource discovery waits for reload. For the remainder of the setup turn, read newly effective context directly after verification instead of claiming that another `before_agent_start` invocation already occurred.