fix: feat: implement evidence/resources and evidence/protocol logging (#1059)
- Add evidence/resources/ and evidence/protocol/ directories with .gitkeep - Add schemas for resources/ and protocol/ to evidence/README.md - Create formulas/run-resources.toml (sense formula: disk/RAM/API/CI metrics, daily cron 06:00 UTC, verdict: ok/warn/critical) - Create formulas/run-protocol.toml (sense formula: TVL/fees/positions/ rebalance frequency via LmTotalEth.s.sol + cast, daily cron 07:00 UTC, verdict: healthy/degraded/offline) - Update STATE.md Co-Authored-By: Claude Sonnet 4.6 <noreply@anthropic.com>
This commit is contained in:
johba
parent
7fd1963480
commit
de014e9b13
6 changed files with 458 additions and 1 deletions
2
STATE.md
2
STATE.md
|
|
@ -50,4 +50,4 @@
|
||||||
- [2026-03-15] txnBot AGENTS.md ENVIRONMENT enum is stale (#784)
|
- [2026-03-15] txnBot AGENTS.md ENVIRONMENT enum is stale (#784)
|
||||||
- [2026-03-20] Adoption milestone state ambiguity in MEMORY.md (#1068)
|
- [2026-03-20] Adoption milestone state ambiguity in MEMORY.md (#1068)
|
||||||
- [2026-03-20] OptimizerV3Push3 as IOptimizer always returns bear defaults — integration risk (#1063)
|
- [2026-03-20] OptimizerV3Push3 as IOptimizer always returns bear defaults — integration risk (#1063)
|
||||||
- [2026-03-20] Red-team schema should add candidate_commit field (#1066)
|
- [2026-03-20] implement evidence/resources and evidence/protocol logging (#1059): formulas/run-resources.toml (disk/RAM/API/CI sense formula, daily cron 06:00 UTC) and formulas/run-protocol.toml (TVL/fees/positions/rebalances sense formula, daily cron 07:00 UTC); evidence/resources/ and evidence/protocol/ directories; schemas in evidence/README.md
|
||||||
|
|
|
||||||
|
|
@ -22,6 +22,10 @@ evidence/
|
||||||
YYYY-MM-DD-prNNN.json # per-scenario pass/fail, gate decision
|
YYYY-MM-DD-prNNN.json # per-scenario pass/fail, gate decision
|
||||||
user-test/
|
user-test/
|
||||||
YYYY-MM-DD.json # per-persona reports, screenshot refs, friction points
|
YYYY-MM-DD.json # per-persona reports, screenshot refs, friction points
|
||||||
|
resources/
|
||||||
|
YYYY-MM-DD.json # disk, RAM, API call counts, budget burn, CI queue depth
|
||||||
|
protocol/
|
||||||
|
YYYY-MM-DD.json # TVL, accumulated fees, position count, rebalance frequency
|
||||||
```
|
```
|
||||||
|
|
||||||
## Delivery Pattern
|
## Delivery Pattern
|
||||||
|
|
@ -215,3 +219,114 @@ Records a UX evaluation run across simulated personas.
|
||||||
| `personas_total` | integer | Total personas evaluated |
|
| `personas_total` | integer | Total personas evaluated |
|
||||||
| `critical_friction_points` | array of strings | Friction points that blocked task completion |
|
| `critical_friction_points` | array of strings | Friction points that blocked task completion |
|
||||||
| `verdict` | string | `"pass"` if all personas completed, `"fail"` otherwise |
|
| `verdict` | string | `"pass"` if all personas completed, `"fail"` otherwise |
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
## Schema: `resources/YYYY-MM-DD.json`
|
||||||
|
|
||||||
|
Records one infrastructure resource snapshot.
|
||||||
|
|
||||||
|
```json
|
||||||
|
{
|
||||||
|
"date": "YYYY-MM-DD",
|
||||||
|
"disk": {
|
||||||
|
"used_bytes": 85899345920,
|
||||||
|
"total_bytes": 107374182400,
|
||||||
|
"used_pct": 80.0
|
||||||
|
},
|
||||||
|
"ram": {
|
||||||
|
"used_bytes": 3221225472,
|
||||||
|
"total_bytes": 8589934592,
|
||||||
|
"used_pct": 37.5
|
||||||
|
},
|
||||||
|
"api": {
|
||||||
|
"anthropic_calls_24h": 142,
|
||||||
|
"anthropic_budget_usd_used": 4.87,
|
||||||
|
"anthropic_budget_usd_limit": 50.0,
|
||||||
|
"anthropic_budget_pct": 9.7
|
||||||
|
},
|
||||||
|
"ci": {
|
||||||
|
"woodpecker_queue_depth": 2,
|
||||||
|
"woodpecker_running": 1
|
||||||
|
},
|
||||||
|
"staleness_threshold_days": 1,
|
||||||
|
"verdict": "ok" | "warn" | "critical"
|
||||||
|
}
|
||||||
|
```
|
||||||
|
|
||||||
|
| Field | Type | Description |
|
||||||
|
|-------|------|-------------|
|
||||||
|
| `date` | string (ISO) | Date of the snapshot |
|
||||||
|
| `disk.used_bytes` | integer | Bytes used on the primary volume |
|
||||||
|
| `disk.total_bytes` | integer | Total bytes on the primary volume |
|
||||||
|
| `disk.used_pct` | number | Percentage of disk used |
|
||||||
|
| `ram.used_bytes` | integer | Bytes of RAM in use |
|
||||||
|
| `ram.total_bytes` | integer | Total bytes of RAM |
|
||||||
|
| `ram.used_pct` | number | Percentage of RAM used |
|
||||||
|
| `api.anthropic_calls_24h` | integer | Anthropic API calls in the past 24 hours |
|
||||||
|
| `api.anthropic_budget_usd_used` | number | USD spent against the Anthropic budget |
|
||||||
|
| `api.anthropic_budget_usd_limit` | number | Configured Anthropic budget ceiling in USD |
|
||||||
|
| `api.anthropic_budget_pct` | number | Percentage of budget consumed |
|
||||||
|
| `ci.woodpecker_queue_depth` | integer | Number of jobs waiting in the Woodpecker CI queue |
|
||||||
|
| `ci.woodpecker_running` | integer | Number of Woodpecker jobs currently running |
|
||||||
|
| `staleness_threshold_days` | integer | Maximum age in days before this record is considered stale (always 1) |
|
||||||
|
| `verdict` | string | `"ok"` (all metrics normal), `"warn"` (≥80% on any dimension), or `"critical"` (≥95% on any dimension) |
|
||||||
|
|
||||||
|
---
|
||||||
|
|
||||||
|
## Schema: `protocol/YYYY-MM-DD.json`
|
||||||
|
|
||||||
|
Records one on-chain protocol health snapshot.
|
||||||
|
|
||||||
|
```json
|
||||||
|
{
|
||||||
|
"date": "YYYY-MM-DD",
|
||||||
|
"block_number": 24500000,
|
||||||
|
"tvl_eth": "1234567890000000000000",
|
||||||
|
"tvl_eth_formatted": "1234.57",
|
||||||
|
"accumulated_fees_eth": "12345678900000000",
|
||||||
|
"accumulated_fees_eth_formatted": "0.012",
|
||||||
|
"position_count": 3,
|
||||||
|
"positions": [
|
||||||
|
{
|
||||||
|
"name": "floor",
|
||||||
|
"tick_lower": -887272,
|
||||||
|
"tick_upper": -200000,
|
||||||
|
"liquidity": "987654321000000000"
|
||||||
|
},
|
||||||
|
{
|
||||||
|
"name": "anchor",
|
||||||
|
"tick_lower": -200000,
|
||||||
|
"tick_upper": 0
|
||||||
|
},
|
||||||
|
{
|
||||||
|
"name": "discovery",
|
||||||
|
"tick_lower": 0,
|
||||||
|
"tick_upper": 887272
|
||||||
|
}
|
||||||
|
],
|
||||||
|
"rebalance_count_24h": 4,
|
||||||
|
"last_rebalance_block": 24499800,
|
||||||
|
"staleness_threshold_days": 1,
|
||||||
|
"verdict": "healthy" | "degraded" | "offline"
|
||||||
|
}
|
||||||
|
```
|
||||||
|
|
||||||
|
| Field | Type | Description |
|
||||||
|
|-------|------|-------------|
|
||||||
|
| `date` | string (ISO) | Date of the snapshot |
|
||||||
|
| `block_number` | integer | Block number at time of snapshot |
|
||||||
|
| `tvl_eth` | string (wei) | Total value locked across all LM positions in wei |
|
||||||
|
| `tvl_eth_formatted` | string | TVL formatted in ETH (2 dp) |
|
||||||
|
| `accumulated_fees_eth` | string (wei) | Fees accumulated by the LiquidityManager in wei |
|
||||||
|
| `accumulated_fees_eth_formatted` | string | Fees formatted in ETH (3 dp) |
|
||||||
|
| `position_count` | integer | Number of active Uniswap V3 positions (expected: 3) |
|
||||||
|
| `positions` | array | One entry per active position |
|
||||||
|
| `positions[].name` | string | Position label: `"floor"`, `"anchor"`, or `"discovery"` |
|
||||||
|
| `positions[].tick_lower` | integer | Lower tick boundary |
|
||||||
|
| `positions[].tick_upper` | integer | Upper tick boundary |
|
||||||
|
| `positions[].liquidity` | string | Liquidity amount in the position (wei-scale integer) |
|
||||||
|
| `rebalance_count_24h` | integer | Number of `recenter()` calls in the past 24 hours |
|
||||||
|
| `last_rebalance_block` | integer | Block number of the most recent `recenter()` call |
|
||||||
|
| `staleness_threshold_days` | integer | Maximum age in days before this record is considered stale (always 1) |
|
||||||
|
| `verdict` | string | `"healthy"` (positions active, TVL > 0), `"degraded"` (position_count < 3 or rebalance stalled), or `"offline"` (TVL = 0 or contract unreachable) |
|
||||||
|
|
|
||||||
0
evidence/protocol/.gitkeep
Normal file
0
evidence/protocol/.gitkeep
Normal file
0
evidence/resources/.gitkeep
Normal file
0
evidence/resources/.gitkeep
Normal file
187
formulas/run-protocol.toml
Normal file
187
formulas/run-protocol.toml
Normal file
|
|
@ -0,0 +1,187 @@
|
||||||
|
# formulas/run-protocol.toml
|
||||||
|
#
|
||||||
|
# On-chain protocol health snapshot — collect TVL, accumulated fees,
|
||||||
|
# position count, and rebalance frequency from the deployed LiquidityManager.
|
||||||
|
# Write a structured JSON evidence file for planner and predictor consumption.
|
||||||
|
#
|
||||||
|
# Type: sense. Read-only — produces metrics only, no git artifacts.
|
||||||
|
#
|
||||||
|
# Staleness threshold: 1 day (matches evidence/protocol/ schema).
|
||||||
|
# Cron: daily at 07:00 UTC (staggered 1 h after run-resources).
|
||||||
|
|
||||||
|
[formula]
|
||||||
|
id = "run-protocol"
|
||||||
|
name = "On-Chain Protocol Health Snapshot"
|
||||||
|
description = "Collect TVL, accumulated fees, position count, and rebalance frequency from the deployed LiquidityManager; write evidence/protocol/{date}.json."
|
||||||
|
type = "sense"
|
||||||
|
# "sense" → read-only, produces metrics only
|
||||||
|
# "act" → produces git artifacts (cf. run-evolution, run-red-team)
|
||||||
|
|
||||||
|
# ── Cron ───────────────────────────────────────────────────────────────────────
|
||||||
|
|
||||||
|
[cron]
|
||||||
|
schedule = "0 7 * * *" # daily at 07:00 UTC (1 h after run-resources)
|
||||||
|
description = "Matches 1-day staleness threshold — one snapshot per day keeps the record fresh."
|
||||||
|
|
||||||
|
# ── Inputs ─────────────────────────────────────────────────────────────────────
|
||||||
|
|
||||||
|
[inputs.rpc_url]
|
||||||
|
type = "string"
|
||||||
|
required = true
|
||||||
|
description = """
|
||||||
|
Base network RPC endpoint used to query on-chain state.
|
||||||
|
Example: https://mainnet.base.org or a running Anvil fork URL.
|
||||||
|
"""
|
||||||
|
|
||||||
|
[inputs.deployments_file]
|
||||||
|
type = "string"
|
||||||
|
required = false
|
||||||
|
default = "onchain/deployments-local.json"
|
||||||
|
description = """
|
||||||
|
Path to the deployments JSON file containing contract addresses.
|
||||||
|
The formula reads LiquidityManager address from this file.
|
||||||
|
Use onchain/deployments.json for mainnet; onchain/deployments-local.json
|
||||||
|
for a local Anvil fork.
|
||||||
|
"""
|
||||||
|
|
||||||
|
[inputs.lookback_blocks]
|
||||||
|
type = "integer"
|
||||||
|
required = false
|
||||||
|
default = 7200
|
||||||
|
description = """
|
||||||
|
Number of blocks to scan for Recenter events when computing
|
||||||
|
rebalance_count_24h (~24 h of Base blocks at ~2 s/block).
|
||||||
|
"""
|
||||||
|
|
||||||
|
# ── Execution ──────────────────────────────────────────────────────────────────
|
||||||
|
|
||||||
|
[execution]
|
||||||
|
script = "scripts/harb-evaluator/run-protocol.sh"
|
||||||
|
invocation = "RPC_URL={rpc_url} DEPLOYMENTS_FILE={deployments_file} LOOKBACK_BLOCKS={lookback_blocks} bash scripts/harb-evaluator/run-protocol.sh"
|
||||||
|
|
||||||
|
# Exit codes:
|
||||||
|
# 0 snapshot written successfully
|
||||||
|
# 2 infrastructure error (RPC unreachable, missing deployments file, forge unavailable, etc.)
|
||||||
|
|
||||||
|
# ── Steps ──────────────────────────────────────────────────────────────────────
|
||||||
|
|
||||||
|
[[steps]]
|
||||||
|
id = "read-addresses"
|
||||||
|
description = """
|
||||||
|
Read the LiquidityManager contract address from {deployments_file}.
|
||||||
|
Fail with exit code 2 if the file is absent or the address is missing.
|
||||||
|
"""
|
||||||
|
|
||||||
|
[[steps]]
|
||||||
|
id = "collect-tvl"
|
||||||
|
description = """
|
||||||
|
Query LiquidityManager total ETH via forge script LmTotalEth.s.sol
|
||||||
|
against {rpc_url}.
|
||||||
|
Records tvl_eth (wei string) and tvl_eth_formatted (ETH, 2 dp).
|
||||||
|
LmTotalEth.s.sol uses exact Uniswap V3 integer math (LiquidityAmounts +
|
||||||
|
TickMath) to sum free ETH, free WETH, and ETH locked across all three
|
||||||
|
positions (floor, anchor, discovery).
|
||||||
|
"""
|
||||||
|
forge_script = "onchain/script/LmTotalEth.s.sol"
|
||||||
|
|
||||||
|
[[steps]]
|
||||||
|
id = "collect-fees"
|
||||||
|
description = """
|
||||||
|
Query accumulated protocol fees from the LiquidityManager via cast call:
|
||||||
|
LiquidityManager.accumulatedFees() → uint256
|
||||||
|
Records accumulated_fees_eth (wei string) and accumulated_fees_eth_formatted
|
||||||
|
(ETH, 3 dp).
|
||||||
|
Falls back to 0 gracefully if the function is not present on the deployed
|
||||||
|
contract (older deployment without fee tracking).
|
||||||
|
"""
|
||||||
|
|
||||||
|
[[steps]]
|
||||||
|
id = "collect-positions"
|
||||||
|
description = """
|
||||||
|
Query the three Uniswap V3 positions held by the LiquidityManager:
|
||||||
|
LiquidityManager.floorPosition() → (tickLower, tickUpper, liquidity)
|
||||||
|
LiquidityManager.anchorPosition() → (tickLower, tickUpper, liquidity)
|
||||||
|
LiquidityManager.discoveryPosition() → (tickLower, tickUpper, liquidity)
|
||||||
|
Records position_count (number of positions with liquidity > 0) and the
|
||||||
|
positions array.
|
||||||
|
"""
|
||||||
|
|
||||||
|
[[steps]]
|
||||||
|
id = "collect-rebalances"
|
||||||
|
description = """
|
||||||
|
Count Recenter events emitted by the LiquidityManager in the past
|
||||||
|
{lookback_blocks} blocks via eth_getLogs.
|
||||||
|
Records:
|
||||||
|
- rebalance_count_24h: total Recenter event count in the window.
|
||||||
|
- last_rebalance_block: block number of the most recent Recenter event
|
||||||
|
(0 if none found in the window).
|
||||||
|
"""
|
||||||
|
event_signature = "Recenter(int24,int24,int24)"
|
||||||
|
|
||||||
|
[[steps]]
|
||||||
|
id = "collect"
|
||||||
|
description = """
|
||||||
|
Assemble all collected metrics into evidence/protocol/{date}.json.
|
||||||
|
Compute verdict:
|
||||||
|
- "offline" if tvl_eth = 0 or RPC was unreachable.
|
||||||
|
- "degraded" if position_count < 3, or rebalance_count_24h = 0 and the
|
||||||
|
protocol has been live for > 1 day.
|
||||||
|
- "healthy" otherwise.
|
||||||
|
Write the file conforming to the schema in evidence/README.md
|
||||||
|
## Schema: protocol/YYYY-MM-DD.json.
|
||||||
|
"""
|
||||||
|
output = "evidence/protocol/{date}.json"
|
||||||
|
schema = "evidence/README.md" # see ## Schema: protocol/YYYY-MM-DD.json
|
||||||
|
|
||||||
|
[[steps]]
|
||||||
|
id = "deliver"
|
||||||
|
description = """
|
||||||
|
Commit evidence/protocol/{date}.json to main.
|
||||||
|
Post a one-line summary comment to the originating issue (if any):
|
||||||
|
verdict, tvl_eth_formatted, accumulated_fees_eth_formatted,
|
||||||
|
position_count, rebalance_count_24h.
|
||||||
|
On "degraded" or "offline": highlight the failing dimension and its value.
|
||||||
|
"""
|
||||||
|
|
||||||
|
# ── Products ───────────────────────────────────────────────────────────────────
|
||||||
|
|
||||||
|
[products.evidence_file]
|
||||||
|
path = "evidence/protocol/{date}.json"
|
||||||
|
delivery = "commit to main"
|
||||||
|
schema = "evidence/README.md" # see ## Schema: protocol/YYYY-MM-DD.json
|
||||||
|
|
||||||
|
[products.issue_comment]
|
||||||
|
delivery = "post to originating issue (if any)"
|
||||||
|
content = "verdict, tvl_eth_formatted, accumulated_fees_eth_formatted, position_count, rebalance_count_24h"
|
||||||
|
on_degraded = "highlight failing dimension and its current value"
|
||||||
|
|
||||||
|
# ── Resources ──────────────────────────────────────────────────────────────────
|
||||||
|
|
||||||
|
[resources]
|
||||||
|
profile = "light"
|
||||||
|
compute = "local — forge script + cast calls only; no Anvil or Docker startup required"
|
||||||
|
rpc = "Base network RPC ({rpc_url}) — read-only calls"
|
||||||
|
concurrency = "safe to run in parallel with other formulas"
|
||||||
|
|
||||||
|
# ── Notes ──────────────────────────────────────────────────────────────────────
|
||||||
|
|
||||||
|
[notes]
|
||||||
|
tvl_metric = """
|
||||||
|
TVL is measured as LiquidityManager total ETH: free ETH + free WETH + ETH
|
||||||
|
locked across all three Uniswap V3 positions (floor, anchor, discovery).
|
||||||
|
Uses the same LmTotalEth.s.sol forge script as run-red-team to ensure
|
||||||
|
consistent measurement methodology.
|
||||||
|
"""
|
||||||
|
|
||||||
|
rebalance_staleness = """
|
||||||
|
A zero rebalance_count_24h on an established deployment indicates the
|
||||||
|
recenter() upkeep bot (services/txnBot) has stalled. The "degraded"
|
||||||
|
verdict triggers a planner alert. On a fresh deployment (< 1 day old)
|
||||||
|
zero rebalances is expected and does not trigger degraded.
|
||||||
|
"""
|
||||||
|
|
||||||
|
fees_fallback = """
|
||||||
|
accumulated_fees_eth falls back to 0 for deployments without fee tracking.
|
||||||
|
The verdict is not affected by a zero fee value alone — only TVL and
|
||||||
|
position_count drive the verdict.
|
||||||
|
"""
|
||||||
155
formulas/run-resources.toml
Normal file
155
formulas/run-resources.toml
Normal file
|
|
@ -0,0 +1,155 @@
|
||||||
|
# formulas/run-resources.toml
|
||||||
|
#
|
||||||
|
# Infrastructure resource snapshot — collect disk usage, RAM trends,
|
||||||
|
# Anthropic API call counts and budget burn, and Woodpecker CI queue depth.
|
||||||
|
# Write a structured JSON evidence file for planner and predictor consumption.
|
||||||
|
#
|
||||||
|
# Type: sense. Read-only — produces metrics only, no git artifacts.
|
||||||
|
#
|
||||||
|
# Staleness threshold: 1 day (matches evidence/resources/ schema).
|
||||||
|
# Cron: daily at 06:00 UTC.
|
||||||
|
|
||||||
|
[formula]
|
||||||
|
id = "run-resources"
|
||||||
|
name = "Infrastructure Resource Snapshot"
|
||||||
|
description = "Collect disk, RAM, API usage, Anthropic budget burn, and CI queue depth; write evidence/resources/{date}.json."
|
||||||
|
type = "sense"
|
||||||
|
# "sense" → read-only, produces metrics only
|
||||||
|
# "act" → produces git artifacts (cf. run-evolution, run-red-team)
|
||||||
|
|
||||||
|
# ── Cron ───────────────────────────────────────────────────────────────────────
|
||||||
|
|
||||||
|
[cron]
|
||||||
|
schedule = "0 6 * * *" # daily at 06:00 UTC
|
||||||
|
description = "Matches 1-day staleness threshold — one snapshot per day keeps the record fresh."
|
||||||
|
|
||||||
|
# ── Inputs ─────────────────────────────────────────────────────────────────────
|
||||||
|
|
||||||
|
[inputs.disk_path]
|
||||||
|
type = "string"
|
||||||
|
required = false
|
||||||
|
default = "/"
|
||||||
|
description = "Filesystem path to measure disk usage for (passed to df)."
|
||||||
|
|
||||||
|
[inputs.anthropic_budget_usd_limit]
|
||||||
|
type = "number"
|
||||||
|
required = false
|
||||||
|
default = 50.0
|
||||||
|
description = "Configured Anthropic budget ceiling in USD. Used to compute budget_pct in the evidence record."
|
||||||
|
|
||||||
|
[inputs.woodpecker_api_url]
|
||||||
|
type = "string"
|
||||||
|
required = false
|
||||||
|
default = "http://localhost:8090"
|
||||||
|
description = "Base URL of the Woodpecker CI API. Set to empty string to skip CI metrics."
|
||||||
|
|
||||||
|
# ── Execution ──────────────────────────────────────────────────────────────────
|
||||||
|
|
||||||
|
[execution]
|
||||||
|
script = "scripts/harb-evaluator/run-resources.sh"
|
||||||
|
invocation = "DISK_PATH={disk_path} ANTHROPIC_BUDGET_USD_LIMIT={anthropic_budget_usd_limit} WOODPECKER_API_URL={woodpecker_api_url} bash scripts/harb-evaluator/run-resources.sh"
|
||||||
|
|
||||||
|
# Exit codes:
|
||||||
|
# 0 snapshot written successfully
|
||||||
|
# 2 infrastructure error (disk command unavailable, JSON write failed, etc.)
|
||||||
|
|
||||||
|
# ── Steps ──────────────────────────────────────────────────────────────────────
|
||||||
|
|
||||||
|
[[steps]]
|
||||||
|
id = "collect-disk"
|
||||||
|
description = """
|
||||||
|
Measure disk usage on {disk_path} via `df -B1 {disk_path}`.
|
||||||
|
Extract used_bytes, total_bytes, and used_pct.
|
||||||
|
"""
|
||||||
|
|
||||||
|
[[steps]]
|
||||||
|
id = "collect-ram"
|
||||||
|
description = """
|
||||||
|
Measure RAM usage via `free -b` (Linux) or `vm_stat` (macOS).
|
||||||
|
Extract used_bytes, total_bytes, and used_pct.
|
||||||
|
"""
|
||||||
|
|
||||||
|
[[steps]]
|
||||||
|
id = "collect-api"
|
||||||
|
description = """
|
||||||
|
Collect Anthropic API metrics:
|
||||||
|
- anthropic_calls_24h: count of API calls in the past 24 hours (read from
|
||||||
|
tmp/anthropic-call-log.jsonl if present; 0 if absent).
|
||||||
|
- anthropic_budget_usd_used: sum of cost_usd entries in the call log for
|
||||||
|
the current calendar day (UTC); 0 if log absent.
|
||||||
|
- anthropic_budget_usd_limit: from {anthropic_budget_usd_limit} input.
|
||||||
|
- anthropic_budget_pct: used / limit * 100 (0 if limit = 0).
|
||||||
|
"""
|
||||||
|
call_log = "tmp/anthropic-call-log.jsonl"
|
||||||
|
|
||||||
|
[[steps]]
|
||||||
|
id = "collect-ci"
|
||||||
|
description = """
|
||||||
|
Query Woodpecker CI API for queue state.
|
||||||
|
GET {woodpecker_api_url}/api/queue/info:
|
||||||
|
- woodpecker_queue_depth: length of the waiting queue.
|
||||||
|
- woodpecker_running: count of currently running jobs.
|
||||||
|
Skipped gracefully (fields set to null) when {woodpecker_api_url} is empty
|
||||||
|
or the endpoint is unreachable.
|
||||||
|
"""
|
||||||
|
|
||||||
|
[[steps]]
|
||||||
|
id = "collect"
|
||||||
|
description = """
|
||||||
|
Assemble all collected metrics into evidence/resources/{date}.json.
|
||||||
|
Compute verdict:
|
||||||
|
- "critical" if disk_used_pct ≥ 95, ram_used_pct ≥ 95,
|
||||||
|
or anthropic_budget_pct ≥ 95.
|
||||||
|
- "warn" if disk_used_pct ≥ 80, ram_used_pct ≥ 80,
|
||||||
|
or anthropic_budget_pct ≥ 80.
|
||||||
|
- "ok" otherwise.
|
||||||
|
Write the file conforming to the schema in evidence/README.md
|
||||||
|
## Schema: resources/YYYY-MM-DD.json.
|
||||||
|
"""
|
||||||
|
output = "evidence/resources/{date}.json"
|
||||||
|
schema = "evidence/README.md" # see ## Schema: resources/YYYY-MM-DD.json
|
||||||
|
|
||||||
|
[[steps]]
|
||||||
|
id = "deliver"
|
||||||
|
description = """
|
||||||
|
Commit evidence/resources/{date}.json to main.
|
||||||
|
Post a one-line summary comment to the originating issue (if any):
|
||||||
|
verdict, disk_used_pct, ram_used_pct, anthropic_budget_pct, ci queue depth.
|
||||||
|
On "warn" or "critical": highlight the breaching dimensions.
|
||||||
|
"""
|
||||||
|
|
||||||
|
# ── Products ───────────────────────────────────────────────────────────────────
|
||||||
|
|
||||||
|
[products.evidence_file]
|
||||||
|
path = "evidence/resources/{date}.json"
|
||||||
|
delivery = "commit to main"
|
||||||
|
schema = "evidence/README.md" # see ## Schema: resources/YYYY-MM-DD.json
|
||||||
|
|
||||||
|
[products.issue_comment]
|
||||||
|
delivery = "post to originating issue (if any)"
|
||||||
|
content = "verdict, disk_used_pct, ram_used_pct, anthropic_budget_pct, ci queue depth"
|
||||||
|
on_warn = "highlight breaching dimensions and current values"
|
||||||
|
|
||||||
|
# ── Resources ──────────────────────────────────────────────────────────────────
|
||||||
|
|
||||||
|
[resources]
|
||||||
|
profile = "light"
|
||||||
|
compute = "local — shell commands only (df, free, curl); no Docker or Anvil required"
|
||||||
|
concurrency = "safe to run in parallel with other formulas"
|
||||||
|
|
||||||
|
# ── Notes ──────────────────────────────────────────────────────────────────────
|
||||||
|
|
||||||
|
[notes]
|
||||||
|
call_log = """
|
||||||
|
tmp/anthropic-call-log.jsonl is expected to have one JSON object per line,
|
||||||
|
each with at minimum:
|
||||||
|
{ "ts": "<ISO timestamp>", "cost_usd": <number> }
|
||||||
|
The file is written by the dark-factory agent loop. When absent the API
|
||||||
|
metrics default to 0 — the snapshot is still written rather than failing.
|
||||||
|
"""
|
||||||
|
|
||||||
|
disk_warn = """
|
||||||
|
Planner MEMORY.md (2026-03-20) notes disk at 79%. The "warn" threshold
|
||||||
|
(≥80%) will fire on the first run-resources pass. Monitor trajectory;
|
||||||
|
evidence pipeline data accumulation will increase disk pressure.
|
||||||
|
"""
|
||||||
Loading…
Add table
Add a link
Reference in a new issue