Skip to content

akiiwa-games/AgentCompass / skills/agent-compass

agent-compass

Flagged

Choose owner, model and effort, outbound authorization, advisor use, and continuation for delegated work. Reads local usage and effective account policy, profiles the task with the configured System-One scorer, then allocates a compatible worker. Use for independent worker or advisor work; ordinary local execution needs no routing.

FileKindSize
SKILL.mdskill md5.5 KB
agents/openai.yamlconfig427 B
calibration/README.mdmarkdown55.2 KB
calibration/arena_capture.pyscript14.0 KB
calibration/community.pyscript25.2 KB
calibration/five_axis.pyscript196.4 KB
calibration/measurements.pyscript68.4 KB
calibration/refresh.pyscript48.6 KB
calibration/series.pyscript10.2 KB
calibration/sources.pyscript24.1 KB
calibration/tiers.pyscript14.6 KB
grok-bot/.gitignoreother47 B
grok-bot/dependency_patch.pyscript4.5 KB
grok-bot/install_cli.pyscript3.6 KB
grok-bot/test-recovery.mjsscript8.6 KB
grok-bot/test_install_cli.pyscript4.1 KB
grok-bot/upstream.jsonconfig3.4 KB
references/accounts.jsonconfig155.9 KB
references/classifier-cases.jsonconfig21.5 KB
references/dispatch-lifecycle.mdmarkdown28.9 KB
references/dispatch-rules.mdmarkdown11.3 KB
references/effort-matching.mdmarkdown5.1 KB
references/fast-modes.mdmarkdown3.5 KB
references/five-axis-scores.jsonconfig3.1 MB
references/gemini-decisions-api.mdmarkdown2.9 KB
references/grok-bot.mdmarkdown13.0 KB
references/herdr.mdmarkdown5.1 KB
references/jev-api.mdmarkdown1.6 KB
references/level-3-conversion.jsonconfig2.4 KB
references/model-capabilities.jsonconfig3.6 KB
references/model-capabilities.mdmarkdown14.6 KB
references/parent-address-gate.mdmarkdown10.1 KB
references/plan-templates.jsonconfig13.2 KB
references/provider-discovery.mdmarkdown11.5 KB
references/provider-usage.mdmarkdown4.1 KB
references/public-data.jsonconfig816 B
references/public-measurements.jsonconfig1.3 MB
references/roll-reference.mdmarkdown26.7 KB
references/router-internals.mdmarkdown93.7 KB
references/task-classifier.mdmarkdown6.1 KB
references/user-scores.jsonconfig6.8 KB
references/work-kinds.mdmarkdown4.7 KB
references/worker-communication.mdmarkdown2.1 KB
scripts/account_fleet.pyscript32.9 KB
scripts/account_state.pyscript22.2 KB
scripts/accounts_override.pyscript14.8 KB
scripts/allocator.pyscript19.8 KB
scripts/antigravity_plan.pyscript18.7 KB
scripts/antigravity_telemetry.pyscript5.1 KB
scripts/auto_bundles.pyscript31.3 KB
scripts/axis_scores.pyscript40.3 KB
scripts/caller_input.pyscript3.0 KB
scripts/candidate_eval.pyscript54.0 KB
scripts/candidate_ranking.pyscript9.5 KB
scripts/capability_catalog.pyscript8.0 KB
scripts/check_router_contract.pyscript2.8 KB
scripts/conftest.pyscript2.5 KB
scripts/cwd_policy.pyscript1.5 KB
scripts/data_age.pyscript1.5 KB
scripts/decision_ledger.pyscript18.9 KB
scripts/demo_plan.pyscript6.3 KB
scripts/dispatch_fallback.pyscript40.3 KB
scripts/dispatch_flow.pyscript58.7 KB
scripts/dispatch_ground_truth.pyscript43.5 KB
scripts/dispatch_registry.pyscript113.7 KB
scripts/entry.pyscript1.7 KB
scripts/env_vars.pyscript580 B
scripts/fast_modes.pyscript2.5 KB
scripts/fixtures/grok-api-responses.jsonconfig907 B
scripts/fixtures/reference-accounts.jsonconfig26.9 KB
scripts/grok_bot_paths.pyscript1005 B
scripts/host_herdr.pyscript16.8 KB
scripts/host_placement.pyscript9.4 KB
scripts/host_policy.pyscript10.1 KB
scripts/launch_failures.pyscript5.9 KB
scripts/launch_values.pyscript12.3 KB
scripts/ledger_snapshot.pyscript11.1 KB
scripts/live_models.pyscript21.1 KB
scripts/model_latch.pyscript7.6 KB
scripts/model_resolution.pyscript18.0 KB
scripts/override_store.pyscript2.2 KB
scripts/plan_state_residue.pyscript5.0 KB
scripts/provider_inventory.pyscript128.2 KB
scripts/provider_retry.pyscript34.6 KB
scripts/public_data.pyscript59.2 KB
scripts/quota_economy.pyscript38.0 KB
scripts/quota_exhaustion.pyscript5.9 KB
scripts/quota_overrides.pyscript25.6 KB
scripts/quota_pools.pyscript29.8 KB
scripts/reference_provenance.pyscript32.3 KB
scripts/released_names.pyscript3.8 KB
scripts/roll.pyscript106.3 KB
scripts/route_judge.pyscript83.4 KB
scripts/route_simulation.pyscript26.3 KB
scripts/router_test_support.pyscript6.7 KB
scripts/run_grok_bot.pyscript18.8 KB
scripts/scorer_client.pyscript55.2 KB
scripts/stage_help.pyscript5.8 KB
scripts/state_upgrade.pyscript24.0 KB
scripts/subscription_cli.pyscript3.6 KB
scripts/task_intake.pyscript33.6 KB
scripts/test_account_identity.pyscript7.1 KB
scripts/test_account_state.pyscript63.3 KB
scripts/test_accounts_override.pyscript30.1 KB
scripts/test_allocator.pyscript94.2 KB
scripts/test_antigravity_plan.pyscript13.2 KB
scripts/test_applicable_quota.pyscript9.1 KB
scripts/test_auto_bundles.pyscript10.8 KB
scripts/test_billing_routing_policy.pyscript9.9 KB
scripts/test_cadence_count.pyscript32.2 KB
scripts/test_calibration_units.pyscript27.0 KB
scripts/test_capability_catalog.pyscript5.0 KB
scripts/test_decision_ledger.pyscript8.5 KB
scripts/test_demo_plan.pyscript3.4 KB
scripts/test_derived_state.pyscript21.6 KB
scripts/test_dispatch.pyscript322.3 KB
scripts/test_dispatch_answers.pyscript14.8 KB
scripts/test_dispatch_fallback.pyscript53.2 KB
scripts/test_entry.pyscript1.6 KB
scripts/test_fast_modes.pyscript9.9 KB
scripts/test_five_axis.pyscript290.1 KB
scripts/test_five_axis_economy.pyscript105.0 KB
scripts/test_fleet_parity.pyscript15.6 KB
scripts/test_grok_bot.pyscript47.7 KB
scripts/test_ground_truth_recording.pyscript68.8 KB
scripts/test_herdr.pyscript14.1 KB
scripts/test_host_policy.pyscript24.6 KB
scripts/test_launch_values.pyscript8.0 KB
scripts/test_metered_usd.pyscript60.4 KB
scripts/test_model_resolution.pyscript9.2 KB
scripts/test_named_route_facts.pyscript28.6 KB
scripts/test_native_hosts.pyscript6.0 KB
scripts/test_new_user_fleet.pyscript3.3 KB
scripts/test_no_hardcode_classification.pyscript75.1 KB
scripts/test_no_legacy_scoring.pyscript6.5 KB
scripts/test_override_store.pyscript2.3 KB
scripts/test_plan_says_why.pyscript44.9 KB
scripts/test_provider_inventory.pyscript83.2 KB
scripts/test_provider_names.pyscript126.0 KB
scripts/test_provider_pools.pyscript14.2 KB
scripts/test_public_data.pyscript93.5 KB
scripts/test_quota_exhaustion.pyscript5.6 KB
scripts/test_rate_limit_windows.pyscript6.3 KB
scripts/test_record_says_why.pyscript26.5 KB
scripts/test_reference_provenance.pyscript16.7 KB
scripts/test_refresh_entry.pyscript110.7 KB
scripts/test_released_names.pyscript4.1 KB
scripts/test_reset_priority.pyscript4.8 KB
scripts/test_roll_answers.pyscript44.7 KB
scripts/test_roll_notifications.pyscript10.4 KB
scripts/test_route_judge.pyscript99.8 KB
scripts/test_route_simulation.pyscript7.0 KB
scripts/test_routing_parity.pyscript32.3 KB
scripts/test_routing_review.pyscript25.1 KB
scripts/test_setup_discovery_parity.pyscript9.5 KB
scripts/test_stage_help.pyscript2.0 KB
scripts/test_state_upgrade.pyscript13.8 KB
scripts/test_turn_refusal.pyscript16.7 KB
scripts/test_usage_cache.pyscript4.9 KB
scripts/test_usage_collector.pyscript21.4 KB
scripts/test_usage_core_claude_accounts.pyscript1.8 KB
scripts/test_usage_core_grok_cli_billing.pyscript15.6 KB
scripts/test_usage_core_magpie.pyscript5.4 KB
scripts/test_usage_core_magpie_pool.pyscript10.8 KB
scripts/test_usage_core_safety.pyscript24.4 KB
scripts/test_user_scores.pyscript102.3 KB
scripts/test_venice_usage.pyscript1.3 KB
scripts/test_window_scope.pyscript5.5 KB
scripts/turn_refusal.pyscript10.7 KB
scripts/usage_cache.pyscript4.7 KB
scripts/usage_collector.pyscript7.4 KB
scripts/usage_core.pyscript142.0 KB
scripts/usage_magpie.pyscript10.6 KB
scripts/usage_reading.pyscript17.1 KB
scripts/user_scores.pyscript72.8 KB
scripts/venice_usage.pyscript2.6 KB
scripts/window_scope.pyscript3.2 KB
scripts/worker_context.pyscript8.9 KB
scripts/worker_sessions.pyscript10.9 KB

scripts/ledger_snapshot.py

View on GitHub
"""The decision snapshot recorded in the ledger."""

import copy

import axis_scores


SNAPSHOT_FULL, SNAPSHOT_COMPACT, SNAPSHOT_EXCLUDED = 6, 8, 16
SNAPSHOT_CAPACITY_KEYS = ('window', 'remaining', 'after', 'step_pct', 'chain_pct', 'capacity_usd', 'remaining_usd',
                          'expected_usd', 'reset_hours', 'end_hours', 'expires_first',
                          'opportunity_cost', 'line_now_usd', 'used_usd', 'quota_source', 'period_hours', 'elapsed_hours',
                          'priced', 'cost', 'quota_price', 'need_start', 'need_end', 'spend_before_reset',
                          'spend_after_reset', 'stale')
SNAPSHOT_DECISION_KEYS = ('action', 'needs_user_decision', 'margin', 'reason', 'alternatives', 'continuity',
                          'fast', 'fast_reason', 'wait_hours', 'start_after_hours', 'better_if_waiting', 'warnings')


def _round(value, places=4):
    if isinstance(value, bool) or not isinstance(value, float):
        return value
    return round(value, places)


def _snapshot_row(candidate, detail):
    """One candidate as it was scored. `detail` keeps every component, the binding window and the notes; the rest keep
    what the comparison table and radar read. Floats are rounded to 4 places to keep a ledger line small."""
    bundle = {k: v for k, v in (candidate.get('bundle') or {}).items() if v is not None and k != 'launch_model'}
    comps = candidate.get('components') or {}
    excluded = [str(r)[:220] for r in (candidate.get('excluded') or [])[:3]]
    row = {'id': candidate.get('id'), 'bundle': bundle, 'score': _round(candidate.get('score')), 'action': candidate.get('action'),
           'excluded': excluded}
    if candidate.get('estimated_axes'):
        row['estimated_axes'] = list(candidate['estimated_axes'])
    if candidate.get('manual_scores'):
        row['manual_scores'] = {
            axis: {k: copy.deepcopy(v) for k, v in entry.items() if k != 'data_reference'}
            for axis, entry in candidate['manual_scores'].items()
            if isinstance(entry, dict)
        }
    if candidate.get('unscored_named_dispatch'):
        row['unscored_named_dispatch'] = True
    if detail:
        row['components'] = {k: _round(v) for k, v in comps.items()}
        row.update({k: _round(candidate.get(k)) for k in ('expected_pct', 'chain_pct', 'expected_usd', 'chain_usd',
                                                           'expected_hours', 'available_in_hours') if candidate.get(k) is not None})
        cap = candidate.get('capacity')
        if isinstance(cap, dict):
            row['capacity'] = {k: _round(cap[k]) for k in SNAPSHOT_CAPACITY_KEYS if k in cap}
        notes = [str(n)[:200] for n in (candidate.get('notes') or [])[:4]]
        if notes:
            row['notes'] = notes
    elif excluded:
        row['components'] = {}
    else:
        shares = tuple(f'share_{a}' for a in axis_scores.FIVE_AXES)
        keep = ('fit', 'opportunity_cost', *axis_scores.FIVE_AXES, *shares, 'continuity',
                'boost',
                'five_axis_level', 'quota_price', 'need_probability', 'need_probability_end', 'substitute_slack',
                'decided_by', 'earliest_reset_hours', 'reading_at', 'duration_multiplier', 'duration_source')
        # a category score or share of 0 is a real value (the anchor's score, the panel's weakest share), kept
        row['components'] = {k: _round(comps[k]) for k in keep
                             if comps.get(k) is not None and (k in axis_scores.FIVE_AXES or k in shares or comps[k])}
        for k in ('fit', 'opportunity_cost'):
            if k in comps:
                row['components'][k] = _round(comps[k])
    if candidate.get('five_axis_headroom') and not excluded:
        row['five_axis_headroom'] = {a: {k: _round(v) if k != 'model' else v for k, v in h.items()}
                                     for a, h in candidate['five_axis_headroom'].items()}
    econ = candidate.get('economy')
    if isinstance(econ, dict) and not excluded:
        # What the README page reads (round 13): per priced window the candidate's spend, cost and need probability at the
        # start and at the end of that spend, plus the key that separated it from the next row.
        row['economy'] = {'cost': _round(econ.get('cost')), 'need_end_max': _round(econ.get('need_end_max')),
                          'binding_window': econ.get('binding_window'),
                          'decided_by': comps.get('decided_by'),
                          'windows': [{k: _round(v) for k, v in w.items()}
                                      for w in econ.get('windows') or ()]}
        if 'expected_usd' not in row and candidate.get('expected_usd') is not None:
            row['expected_usd'] = _round(candidate['expected_usd'])
        if 'expected_hours' not in row and candidate.get('expected_hours') is not None:
            row['expected_hours'] = _round(candidate['expected_hours'])
        line = candidate_quota_line(candidate)
        if line is not None:
            row['quota_line'] = line
    return row


def candidate_quota_line(candidate):
    """The window this candidate's cost is binding on, as the README quota page reads it (round 16, J1): {window: 'source/window'
    (the key in the snapshot's quota_line.windows), hours: the candidate's hours, spend_usd: its spend of that window}. None
    when the candidate prices no long window. The allocator is the one writer; `economy` and `quota_windows` stay as they are."""
    econ = candidate.get('economy')
    bound = econ.get('binding_window') if isinstance(econ, dict) else None
    if not bound:
        return None
    for w in econ.get('windows') or ():
        if w.get('window') == bound and w.get('quota_source') is not None:
            hours = candidate.get('expected_hours')
            return {'window': f"{w['quota_source']}/{bound}", 'hours': _round(hours if hours is not None else 0.0),
                    'spend_usd': _round(w.get('spend_usd'))}
    return None


def snapshot_quota_line(quota_windows):
    """`snapshot.quota_line` for the README quota page: every priced long window of the decision, keyed by its label, with the
    capacity C, the use U and the clock the page draws (elapsed hours since the period's start, hours to the reset; None =
    never resets). `account` and `window_id` are the usage collector's keys, where the page may look its readings up."""
    windows = {}
    for w in quota_windows or ():
        label = w.get('label')
        if not label:
            continue
        windows[label] = {'label': label, 'capacity_usd': w.get('capacity_usd'), 'used_usd': w.get('used_usd'),
                          'elapsed_hours': w.get('elapsed_hours'), 'reset_in_hours': w.get('reset_in_hours'),
                          'account': w.get('quota_source'), 'window_id': w.get('window')}
    return {'windows': windows} if windows else None


def with_quota_line(result):
    """`result` with the quota-line block the quota page reads, derived when the result does not carry one: a fresh plan
    (a simulation or a recomputed replay) or a snapshot recorded without it still holds `decision.system.quota_windows` and
    each candidate's `economy`, and the block is built from them by the same two functions the snapshot uses. A result
    that carries a block, predates per-candidate `economy`, or has no recorded quota windows, comes back unchanged."""
    if not isinstance(result, dict) or isinstance(result.get('quota_line'), dict):
        return result
    eligible = [c for c in (result.get('ranking') or ()) if isinstance(c, dict) and not c.get('excluded')]
    if eligible and not any(isinstance(c.get('economy'), dict) or isinstance(c.get('quota_line'), dict) for c in eligible):
        return result
    block = snapshot_quota_line(((result.get('decision') or {}).get('system') or {}).get('quota_windows'))
    if block is None:
        return result
    ranking = []
    for c in result.get('ranking') or ():
        line = candidate_quota_line(c) if isinstance(c, dict) and 'quota_line' not in c and not c.get('excluded') else None
        ranking.append({**c, 'quota_line': line} if line is not None else c)
    return {**result, 'quota_line': block, 'ranking': ranking}


def decision_snapshot(result, task=None):
    """The decision as the allocator made it, small enough for one ledger line, or None when `result` has no ranking.

    Holds the task demand, the scoring system in force (weights, harness order) and the ranking
    with each candidate's component scores: every component for the best `SNAPSHOT_FULL` eligible candidates, the
    comparison-table fields for the next `SNAPSHOT_COMPACT`, and the first reasons for `SNAPSHOT_EXCLUDED` blocked ones.
    Readers replay this as written and never re-plan the task against later quota or configuration."""
    ranking = result.get('ranking') if isinstance(result, dict) else None
    if not isinstance(ranking, list):
        return None
    decision = result.get('decision') or {}
    eligible = [c for c in ranking if not c.get('excluded') and c.get('score') is not None]
    blocked = [c for c in ranking if c.get('excluded') or c.get('score') is None]
    system = decision.get('system') or {}
    rows = ([_snapshot_row(c, True) for c in eligible[:SNAPSHOT_FULL]]
            + [_snapshot_row(c, False) for c in eligible[SNAPSHOT_FULL:SNAPSHOT_FULL + SNAPSHOT_COMPACT]]
            + [_snapshot_row(c, False) for c in blocked[:SNAPSHOT_EXCLUDED]])
    d_task = decision.get('task') or {}
    task_in = task if isinstance(task, dict) else (result.get('task') if isinstance(result.get('task'), dict) else {})
    raw_profile = d_task.get('profile') if isinstance(d_task.get('profile'), dict) else (
        task_in.get('profile') if isinstance(task_in.get('profile'), dict) else None)
    clean_profile = {k: v for k, v in raw_profile.items() if k != 'reason'} if isinstance(raw_profile, dict) else None
    task_meta = {k: v for k, v in (
        ('objective', (task_in.get('objective') or '')[:200] or None),
        ('work_kind', task_in.get('work_kind') or d_task.get('kind')),
        ('profile', clean_profile),
    ) if v is not None}
    return {
        'v': 1, 'stage': result.get('stage'), 'source': result.get('source'), 'observed_at': result.get('observed_at'),
        'task': task_meta,
        'choice': result.get('choice'),
        'selected_bundle': result.get('selected_bundle'),
        'expected_usd': _round(result.get('expected_usd')),
        **({'quota_line': quota_line} if (quota_line := snapshot_quota_line(system.get('quota_windows'))) else {}),
        'decision': {**{k: decision[k] for k in SNAPSHOT_DECISION_KEYS if decision.get(k) is not None},
                     'task': {k: v for k, v in d_task.items() if k != 'profile'} | (
                         {'profile': clean_profile} if clean_profile is not None else {}),
                     'system': {'weights': system.get('weights'),
                                'quota_windows': system.get('quota_windows')}},
        'ranking': rows,
        'ranking_total': len(ranking), 'ranking_eligible': len(eligible),
        'ranking_omitted': {'eligible': max(0, len(eligible) - SNAPSHOT_FULL - SNAPSHOT_COMPACT),
                            'excluded': max(0, len(blocked) - SNAPSHOT_EXCLUDED)},
    }
reverse-engineer-anythingmorluto/reaReverse engineer native, managed, Electron/JavaScript, packaged, firmware, and browser targets with REA, and build agent environments (AgentEnv bundles) of web applications from a live instance or a capture, and of Linux desktop apps with a task from a request. Use shipped-artifact or requested runtime evidence to explain features, compare versions, decompile code, or guide a reconstruction. Skip REA for ordinary source-repository architecture analysis.Stars: 471kUpdated Claude CodeCodexVerifiedInstall
diagnosing-superpowersobra/superpowersUse when a superpowers session went wrong and your human partner wants to know why — repeated work, ignored plans, stumbles, poor results, a skill that didn't fire, "it took too long", "why is it so expensive", "what is it doing" — or wants to build a bug report for the superpowers maintainers, for the current session or a past one identified by id or path, on any harness.Stars: 322kUpdated Claude CodeCodexVerifiedInstall
dispatching-parallel-agentsobra/superpowersUse when facing 2+ independent tasks that can be worked on without shared state or sequential dependenciesStars: 322kUpdated Claude CodeCodexVerifiedInstall
executing-plansobra/superpowersUse when executing an implementation plan in the current session as the implementer yourself — your human partner chose inline execution, or no subagent tool is availableStars: 322kUpdated Claude CodeCodexVerifiedInstall