📖About

INI-2199 made criteria mandatory but 0107 slipped through because it HAD criteria — 6 of them, all must_have, all met. But they were process-meta checks: Design row exists, shipped via bypass, no Themis. The reviewer couldn't tell they were measuring process, not output. FIX: add an LLM-based quality gate after the existence gate in el_initiative_creator_01. Pass the success criteria list to a lightweight LLM call (via call_claude_raw) asking: are these criteria measuring OUTPUT (what the user can do after the work) or PROCESS (what steps were followed)? If ALL criteria are process-type, reject with the LLM's explanation. If at least one is output-type, allow. The LLM call is fast (~2s with DeepSeek) and fail-open (if the LLM call fails, allow the filing — dont block real work on an infra hiccup).

⚖️Decisions

• Auto-detected remediation target INI-999900440 from title/description scan (matched CAROL-INI-2199 -> row id 999900440 (CAROL-INI-2199-00: Mandatory success criteria gate at filing — refuse to create )); override by setting remediates_initiative_id explicitly at bypass_start. (system-auto-detect)
• Elrond's bypass methodology checklist (a reminder, not a gate -- you've got this): 0. File it requested_mode='bypass' (planner-vs-bypass is a deliberate choice). bypass_start REFUSES a non-bypass initiative (CAROL-INI-1846), and the dispatcher only skips the bypass lane when the mode says bypass -- a 'planner' mistag lets Merlin's pipeline grab the placeholder step and block your finished work. 1. Filed as planned status -- let the bypass claim/activate it; never file active. 2. Open the bypass (bypass_start) with your droid id + the remediation answer (remediates_initiative_id=NNN, or remediates_nothing=True). 3. Work the blocks for your work-type: template -> design -> code -> test -> review. Do the real work; record decisions on the initiative as you make them. 4. Reality is recorded for you at close -- code (files changed), each decision, and the twin-review verdict become real activities tied to this initiative and show in the Activity Tracker like a planner run (CAROL-INI-1840). No dummy rows. 5. Keep the initiative status moving; it parks in 'reviewing' and is tagged uat-pending for you at close (CAROL-INI-1836), so the stuck-watchdog leaves it alone until UAT. 6. Close runs the gates (design/architecture compliance + caller-audit). If a gate flags something pre-existing or unrelated to your change, waive it with a clear written rationale -- audit, don't skip. 7. Bypass skips the planner's auto-orchestration, NOT the standards. Same template checklist, same review, same observability as a planner run. (elrond)
• [status-router] planned -> executing | event=bypass_executing | bypass transition (or-bx-01)
• [status-router] executing -> reviewing | event=bypass_reviewing | bypass transition (or-bx-01)
• Elrond re-scoped success criterion 999900446 (replace) on Albus's prescription — Policy P.01.02.04.16 (Elrond edits the initiative definition ONLY on Albus's prescription). Albus diagnosis: The original criterion 'reject process-only success criteria at filing' is a meta-goal with no measurable target and no implementation path. Replacing it with a bounded, buildable specification for an LLM-based gate gives the initiative a real deliverable that can be verified through a single step. (elrond)
• Elrond re-scoped success criterion 999900446 (replace) on Albus's prescription — Policy P.01.02.04.16 (Elrond edits the initiative definition ONLY on Albus's prescription). Albus diagnosis: The original criterion 'reject process-only success criteria at filing' is a meta-goal with no measurable target and no implementation path. Replacing it with a bounded, buildable specification for an LLM-based gate gives the initiative a real deliverable that can be verified through a single step. (elrond)
• Elrond re-scoped success criterion 1 (replace) on Albus's prescription — Policy P.01.02.04.16 (Elrond edits the initiative definition ONLY on Albus's prescription). Albus diagnosis: The original criterion 'reject process-only success criteria' is self-referentially ambiguous and cannot be enforced by any LLM because 'process-only' is undefined. The replacement grounds the gate in a bounded, testable rule: criteria must reference functional deliverables, not process artifacts. This makes the success criterion measurable and the gate's job unambiguous. (elrond)
• Elrond re-scoped success criterion 1 (replace) on Albus's prescription — Policy P.01.02.04.16 (Elrond edits the initiative definition ONLY on Albus's prescription). Albus diagnosis: The original criterion 'reject process-only success criteria' is self-referentially ambiguous and cannot be enforced by any LLM because 'process-only' is undefined. The replacement grounds the gate in a bounded, testable rule: criteria must reference functional deliverables, not process artifacts. This makes the success criterion measurable and the gate's job unambiguous. (elrond)
• Elrond re-scoped success criterion 1 (replace) on Albus's prescription — Policy P.01.02.04.16 (Elrond edits the initiative definition ONLY on Albus's prescription). Albus diagnosis: The original criterion 'reject process-only success criteria at filing' is impossible to satisfy retroactively for already-filed initiative 0107. Replacing it with a forward-looking gate that operates at filing time creates an achievable, verifiable, and testable goal. (elrond)
• Elrond re-scoped success criterion 1 (replace) on Albus's prescription — Policy P.01.02.04.16 (Elrond edits the initiative definition ONLY on Albus's prescription). Albus diagnosis: The original criterion 'reject process-only success criteria at filing' is impossible to satisfy retroactively for already-filed initiative 0107. Replacing it with a forward-looking gate that operates at filing time creates an achievable, verifiable, and testable goal. (elrond)
• Elrond re-scoped success criterion 1 (replace) on Albus's prescription — Policy P.01.02.04.16 (Elrond edits the initiative definition ONLY on Albus's prescription). Albus diagnosis: The original criterion was unbounded ('reject process-only success criteria at filing') — it cannot be proven exhaustively. The replacement sets a bounded, testable goal: one concrete rejection of one real initiative with demonstrable process-only criteria, with evidence captured in logs. (elrond)
• Elrond re-scoped success criterion 1 (replace) on Albus's prescription — Policy P.01.02.04.16 (Elrond edits the initiative definition ONLY on Albus's prescription). Albus diagnosis: The original criterion was unbounded ('reject process-only success criteria at filing') — it cannot be proven exhaustively. The replacement sets a bounded, testable goal: one concrete rejection of one real initiative with demonstrable process-only criteria, with evidence captured in logs. (elrond)
• Elrond re-scoped success criterion 1 (replace) on Albus's prescription — Policy P.01.02.04.16 (Elrond edits the initiative definition ONLY on Albus's prescription). Albus diagnosis: The original criterion 'LLM-based criteria quality gate' is an infrastructure build, not a review closure. Replacing it with a bounded validation script that can be implemented and tested in a single step makes the initiative completable. The replacement is scoped to a concrete deliverable: a script + one passing test case against real data. (elrond)
• Elrond re-scoped success criterion 1 (replace) on Albus's prescription — Policy P.01.02.04.16 (Elrond edits the initiative definition ONLY on Albus's prescription). Albus diagnosis: The original criterion 'LLM-based criteria quality gate' is an infrastructure build, not a review closure. Replacing it with a bounded validation script that can be implemented and tested in a single step makes the initiative completable. The replacement is scoped to a concrete deliverable: a script + one passing test case against real data. (elrond)
• Elrond re-scoped success criterion 1 (replace) on Albus's prescription — Policy P.01.02.04.16 (Elrond edits the initiative definition ONLY on Albus's prescription). Albus diagnosis: The original criterion 'criteria exist and are must_have' allows any filing with criteria to pass, even if those criteria are process-meta. The gate doesn't evaluate criteria quality — it only checks presence. This is the root cause of the deadlock. (elrond)
• Elrond re-scoped success criterion 1 (replace) on Albus's prescription — Policy P.01.02.04.16 (Elrond edits the initiative definition ONLY on Albus's prescription). Albus diagnosis: The original criterion 'criteria exist and are must_have' allows any filing with criteria to pass, even if those criteria are process-meta. The gate doesn't evaluate criteria quality — it only checks presence. This is the root cause of the deadlock. (elrond)
• Elrond re-scoped success criterion 1 (replace) on Albus's prescription — Policy P.01.02.04.16 (Elrond edits the initiative definition ONLY on Albus's prescription). Albus diagnosis: The current criteria implicitly assume the LLM gate already exists ('reject process-only success criteria at filing') rather than defining a buildable deliverable. The gate itself is what the initiative must create, not what it relies on. Fixed to describe a buildable artifact with concrete success demonstration. (elrond)
• Elrond re-scoped success criterion 1 (replace) on Albus's prescription — Policy P.01.02.04.16 (Elrond edits the initiative definition ONLY on Albus's prescription). Albus diagnosis: The current criteria implicitly assume the LLM gate already exists ('reject process-only success criteria at filing') rather than defining a buildable deliverable. The gate itself is what the initiative must create, not what it relies on. Fixed to describe a buildable artifact with concrete success demonstration. (elrond)
• Elrond re-scoped success criterion 999900446 (replace) on Albus's prescription — Policy P.01.02.04.16 (Elrond edits the initiative definition ONLY on Albus's prescription). Albus diagnosis: The original criterion was too vague — 'reject process-only success criteria at filing' cannot be implemented because the definition of 'process-only' was unclear. The example INI-0107 had 6 must-have criteria, all met, but they were all process-meta checks (existence of design row, bypass flag, no Themis). The corrected criterion bounds 'process-meta' to a concrete pattern set and allows mixed cr (elrond)
• Orion remediated: INI-999900496 bypass closed — CAROL-INI-696 close-marker: the Orion bypass INI-999900496 filed against this parent reached terminal state (closed). This row's literal prefix Orion remediated: is the canonical signal the cookbook-155 dispatcher gate looks for. (shared.bypass.bypass_end)
• [status-router] reviewing -> blocked | event=operator_put | PUT /api/initiatives (operator)
• Orion remediated: INI-999900535 bypass closed — CAROL-INI-696 close-marker: the Orion bypass INI-999900535 filed against this parent reached terminal state (closed). This row's literal prefix Orion remediated: is the canonical signal the cookbook-155 dispatcher gate looks for. (shared.bypass.bypass_end)
• Orion remediated: Albus RSI group diagnosis (via INI 999900490): [procedural, confidence high] The initiative is blocked because the operator manually put it to blocked after a reviewer reopened it for rework, indicating that the execution artifacts did not satisfy the success criteria. The prior diagnosis for a similar initiative (1000107) highlighted that criteria were meta-checks rather than substantive, and although the current criteria appear improved, the pipeline still lacks proper evidence capture automation, causing the operator to intervene when progress stalls. (orion)
• [status-router] blocked -> closed | event=operator_put | PUT /api/initiatives (operator)
• [rsi-group-cure] Cured by the group diagnosis on INI 999900490 (shared cause operator_put); retriggered as INI 999900924. Root cause: [procedural, confidence high] The initiative is blocked because the operator manually put it to blocked after a reviewer reopened it for rework, indicating that the execution artifacts did not satisfy the success criteria. The prior diagnosis for a similar initiative (1000107) highlighted that criteria were meta-checks rather than substantive, and although the current criteria appear improved, the (elrond.rsi_loop)

✅Success criteria

• Filing with process-only criteria (e.g. shipped via bypass, design row exists) is rejected with a clear error explaining which criteria are process-meta and need replacement (must_have)
• Filing with real output criteria (e.g. endpoint returns 200, badge renders correctly) is accepted (must_have)
• The LLM call for classification is fast (<5s) and uses the existing DeepSeek provider (must_have)