@mrjamesakpan Per-set history should help with progression questions. If you want Pennah in HOL Guard, the MCP server guide covers the listing details and PR steps: github.com/hashgraph-online/hol-guard/blob/main/d…
2 likes · 12 views
Code map extension for HOL Guard
Code map extension for HOL Guard
I maintain where-are-we (github.com/ngavrish/where-are-we), a CLI that builds a map of a repository (where things are, what calls what, which tests cover what) so a coding agent can ask the map instead of grepping its way around.
Almost everything an agent does with it is reading. But a few flags change things outside the map:
- --agent-file rewrites a section of CLAUDE.md / AGENTS.md
- --install-hook claude edits ~/.claude/settings.json, so the tool runs at every session start
- --specs / --spec-cmd fetch from a ticket tracker, and --spec-cmd runs a command once per ticket
Showcase
When an agent fires one of these mid-task, you find out later, when your agent instructions or hooks look different and you don't know why. My choices were to approve every call by hand (hundreds of harmless reads) or approve none of them.
So I wrote an extension for HOL Guard (github.com/hashgraph-online/hol-guard), an open-source local firewall that sits between the agent and the shell. It got merged last week. What it does:
- read-only calls (--ask, --callers, rebuilding the map) pass without a prompt
- repo writes and tracker calls pause for your OK
- hook installation is marked high risk
- unknown flags also pause, so a newer version of the tool can't slip past it
What it does NOT catch:
- tracker access configured through env vars (SPEC_ROOTS, SPEC_FETCH_CMD, RUNS_API_READ) instead of flags
- anything that doesn't go through the agent's shell, like a git hook it installed earlier
- it's off by default; you have to enable it in Guard
If you're building CLI tools that agents call, I'm curious how you handle this. Do you split read and write flags yourself, or leave it to the agent harness?
Disclosure: I wrote both where-are-we and this extension. I'm not affiliated with HOL.
Code map extension for HOL Guard
I maintain where-are-we (github.com/ngavrish/where-are-we), a CLI that builds a map of a repository (where things are, what calls what, which tests cover what) so a coding agent can ask the map instead of grepping its way around.
Almost everything an agent does with it is reading. But a few flags change things outside the map:
- --agent-file rewrites a section of CLAUDE.md / AGENTS.md
- --install-hook claude edits ~/.claude/settings.json, so the tool runs at every session start
- --specs / --spec-cmd fetch from a ticket tracker, and --spec-cmd runs a command once per ticket
Showcase
When an agent fires one of these mid-task, you find out later, when your agent instructions or hooks look different and you don't know why. My choices were to approve every call by hand (hundreds of harmless reads) or approve none of them.
So I wrote an extension for HOL Guard (github.com/hashgraph-online/hol-guard), an open-source local firewall that sits between the agent and the shell. It got merged last week. What it does:
- read-only calls (--ask, --callers, rebuilding the map) pass without a prompt
- repo writes and tracker calls pause for your OK
- hook installation is marked high risk
- unknown flags also pause, so a newer version of the tool can't slip past it
What it does NOT catch:
- tracker access configured through env vars (SPEC_ROOTS, SPEC_FETCH_CMD, RUNS_API_READ) instead of flags
- anything that doesn't go through the agent's shell, like a git hook it installed earlier
- it's off by default; you have to enable it in Guard
If you're building CLI tools that agents call, I'm curious how you handle this. Do you split read and write flags yourself, or leave it to the agent harness?
Disclosure: I wrote both where-are-we and this extension. I'm not affiliated with HOL.
1 point · 2 comments
Where AI agents find a data service in 2026: MCP registries, x402 directories and ERC-8004
People find a website through search. AI agents mostly don't browse: they read registries, manifests and machine-readable files, then call what they find. If your service isn't in those places, agents never learn it exists, however good your pages are.
We build Agentic Finance Graph, an independent ledger of the money AI agents hold and spend on Base. Every ERC-8004 registration is joined to the wallet that spends for it, and every stablecoin payment is checked against its receipt. The figures are dated, every one has a definition, and they're free to read.
These are the places we listed it, what each one asks of you, and what we'd tell anyone doing the same.
1. MCP registries: where AI clients look for tools
Our MCP server is remote and stateless. It needs no key, and its 11 tools are all read-only:
https://agenticfinancegraph.com/mcp (Streamable HTTP)
In Claude Code:
claude mcp add --transport http afg https://agenticfinancegraph.com/mcp
In Cursor, Windsurf or VS Code, add a remote MCP server with that URL. Clients that only speak stdio can run the open-source bridge: npx -y github:AgenticFinanceGraph/agentic-finance-graph-mcp mcp.
Where it is listed:
- Official MCP Registry as com.agenticfinancegraph/agentic-finance-graph. The name is verified against our domain through /.well-known/mcp-registry-auth, an Ed25519 public key, so nobody else can publish under it.
- Smithery, under our own namespace.
- Glama. Glama indexes public GitHub repos by itself. When we tried to submit, it answered "already exists", so we claimed the existing page instead.
The tools: market_state, find_agent, agent_record, ranked_agents, detections, definition, metric_history, since_last, agent_statement, counterparty_preview and evidence_pack_quote. None of them can move funds.
2. x402 directories: where agents look for things to pay for
Most of the data is free. On top of it we sell nine evidence packs: dated, citable files about one agent or one address, priced from 1 to 25 cents in USDC on Base, paid over x402 with no account.
Without payment, a pack URL answers 402 Payment Required with the exact terms:
curl -i "https://agenticfinancegraph.com/check/liveness?id=53008"
# HTTP/1.1 402 Payment Required
# PAYMENT-REQUIRED: <base64: x402Version 2, amount 10000 (0.01 USDC), network eip155:8453>
Where it is listed:
- Coinbase's x402 Bazaar (the CDP discovery catalogue). All nine packs are listed. The Bazaar adds a resource only after a real payment has settled, so we bought each pack once ourselves.
- x402scan, which registered all 37 of our paid resources from our OpenAPI file.
- Our own manifest at /.well-known/x402.
One thing we learned the hard way: directories probe bare routes without parameters. A route that answers 400 instead of 402 looks broken to them. Ours answers 402 with the published terms even when no id is given.
3. An agent-readable website
- llms.txt: what we measure, when to call us, which call is free, and which pack costs what.
- openapi.json: every endpoint, including the paid ones.
- /.well-known/agent-card.json: an A2A agent card.
- robots.txt explicitly welcomes retrieval and answer-engine crawlers. Until September it blocked our own API from them, which left AI answers quoting stale prose instead of live figures.
4. An identity: ERC-8004
Agentic Finance Graph is itself registered as ERC-8004 agent #95875 on Base. If we track agents' identities and wallets, we should have one too.
5. Open source on GitHub
- agentic-finance-graph-mcp: the MCP server and CLI (Apache-2.0), with a security scan on every push.
- agentic-finance-graph-spec: the open accounting spec behind every figure (CC BY 4.0).
- Listed in hashgraph-online/awesome-ai-plugins.
What we'd tell anyone listing a service for agents
1. Verify your name. Domain-verified registry names stop impostors and cost one TXT record or one well-known file.
2. Answer bare routes. Catalogues probe endpoints with no parameters. Give them a useful 402, never a confusing 400.
3. Settle once. Some x402 catalogues list a resource only after a real payment. Budget a few cents to buy your own products.
4. Keep a free path. Agents (and people) trust a paid pack more when the free record it packages is one call away.
5. Date and define every number. Agents quote you. Give them an as_of and a definition id to quote with it.
Everything above is free to read under CC BY 4.0: agenticfinancegraph.com. The MCP server, the API and the spec are public. If you build agents that pay, tell us what you'd want to ask the ledger.
@TheInsumerModel mcp-server-insumer gives agents 27 condition-based access tools, signed attestations over 37 chains, via npx. Fits HOL Guard MCP contributions as mcp.insumer.json in contributions/mcp-servers/: add it, open a PR github.com/hashgraph-online/hol-guard/blob/main/C…
3 likes · 21 views
@dwhitedesign cuelume plays a sound when a long Claude Code turn ends so you can look away. Fits awesome-ai-plugins Tools & Integrations: add a one-sentence README entry and open your PR github.com/hashgraph-online/awesome-ai-plugins/bl…
3 likes · 37 views
@_overment Streamable-HTTP MCP server template updated for the latest spec. Fits awesome-ai-plugins Tools & Integrations: add a one-sentence README entry and open your PR github.com/hashgraph-online/awesome-ai-plugins/bl…
3 likes · 25 views
@olanetsoft Agent Lookout shows your running Claude Code and Codex sessions on one local page. Fits awesome-ai-plugins Tools & Integrations: add a one-sentence README entry and open your PR github.com/hashgraph-online/awesome-ai-plugins/bl…
3 likes · 45 views
Mentions
@mrjamesakpan Per-set history should help with progression questions. If you want Pennah in HOL Guard, the MCP server guide covers the listing details and PR steps: github.com/hashgraph-online/hol-guard/blob/main/d…
Code map extension for HOL Guard Code map extension for HOL Guard I maintain where-are-we (github.com/ngavrish/where-are-we), a CLI that builds a map of a repository (where things are, what calls what, which tests cover what) so a coding agent can ask the map instead of grepping its way around. Almost everything an agent does with it is reading. But a few flags change things outside the map: - --agent-file rewrites a section of CLAUDE.md / AGENTS.md - --install-hook claude edits ~/.claude/settings.json, so the tool runs at every session start - --specs / --spec-cmd fetch from a ticket tracker, and --spec-cmd runs a command once per ticket Showcase When an agent fires one of these mid-task, you find out later, when your agent instructions or hooks look different and you don't know why. My choices were to approve every call by hand (hundreds of harmless reads) or approve none of them. So I wrote an extension for HOL Guard (github.com/hashgraph-online/hol-guard), an open-source local firewall that sits between the agent and the shell. It got merged last week. What it does: - read-only calls (--ask, --callers, rebuilding the map) pass without a prompt - repo writes and tracker calls pause for your OK - hook installation is marked high risk - unknown flags also pause, so a newer version of the tool can't slip past it What it does NOT catch: - tracker access configured through env vars (SPEC_ROOTS, SPEC_FETCH_CMD, RUNS_API_READ) instead of flags - anything that doesn't go through the agent's shell, like a git hook it installed earlier - it's off by default; you have to enable it in Guard If you're building CLI tools that agents call, I'm curious how you handle this. Do you split read and write flags yourself, or leave it to the agent harness? Disclosure: I wrote both where-are-we and this extension. I'm not affiliated with HOL.
Code map extension for HOL Guard I maintain where-are-we (github.com/ngavrish/where-are-we), a CLI that builds a map of a repository (where things are, what calls what, which tests cover what) so a coding agent can ask the map instead of grepping its way around. Almost everything an agent does with it is reading. But a few flags change things outside the map: - --agent-file rewrites a section of CLAUDE.md / AGENTS.md - --install-hook claude edits ~/.claude/settings.json, so the tool runs at every session start - --specs / --spec-cmd fetch from a ticket tracker, and --spec-cmd runs a command once per ticket Showcase When an agent fires one of these mid-task, you find out later, when your agent instructions or hooks look different and you don't know why. My choices were to approve every call by hand (hundreds of harmless reads) or approve none of them. So I wrote an extension for HOL Guard (github.com/hashgraph-online/hol-guard), an open-source local firewall that sits between the agent and the shell. It got merged last week. What it does: - read-only calls (--ask, --callers, rebuilding the map) pass without a prompt - repo writes and tracker calls pause for your OK - hook installation is marked high risk - unknown flags also pause, so a newer version of the tool can't slip past it What it does NOT catch: - tracker access configured through env vars (SPEC_ROOTS, SPEC_FETCH_CMD, RUNS_API_READ) instead of flags - anything that doesn't go through the agent's shell, like a git hook it installed earlier - it's off by default; you have to enable it in Guard If you're building CLI tools that agents call, I'm curious how you handle this. Do you split read and write flags yourself, or leave it to the agent harness? Disclosure: I wrote both where-are-we and this extension. I'm not affiliated with HOL.
Where AI agents find a data service in 2026: MCP registries, x402 directories and ERC-8004 People find a website through search. AI agents mostly don't browse: they read registries, manifests and machine-readable files, then call what they find. If your service isn't in those places, agents never learn it exists, however good your pages are. We build Agentic Finance Graph, an independent ledger of the money AI agents hold and spend on Base. Every ERC-8004 registration is joined to the wallet that spends for it, and every stablecoin payment is checked against its receipt. The figures are dated, every one has a definition, and they're free to read. These are the places we listed it, what each one asks of you, and what we'd tell anyone doing the same. 1. MCP registries: where AI clients look for tools Our MCP server is remote and stateless. It needs no key, and its 11 tools are all read-only: https://agenticfinancegraph.com/mcp (Streamable HTTP) In Claude Code: claude mcp add --transport http afg https://agenticfinancegraph.com/mcp In Cursor, Windsurf or VS Code, add a remote MCP server with that URL. Clients that only speak stdio can run the open-source bridge: npx -y github:AgenticFinanceGraph/agentic-finance-graph-mcp mcp. Where it is listed: - Official MCP Registry as com.agenticfinancegraph/agentic-finance-graph. The name is verified against our domain through /.well-known/mcp-registry-auth, an Ed25519 public key, so nobody else can publish under it. - Smithery, under our own namespace. - Glama. Glama indexes public GitHub repos by itself. When we tried to submit, it answered "already exists", so we claimed the existing page instead. The tools: market_state, find_agent, agent_record, ranked_agents, detections, definition, metric_history, since_last, agent_statement, counterparty_preview and evidence_pack_quote. None of them can move funds. 2. x402 directories: where agents look for things to pay for Most of the data is free. On top of it we sell nine evidence packs: dated, citable files about one agent or one address, priced from 1 to 25 cents in USDC on Base, paid over x402 with no account. Without payment, a pack URL answers 402 Payment Required with the exact terms: curl -i "https://agenticfinancegraph.com/check/liveness?id=53008" # HTTP/1.1 402 Payment Required # PAYMENT-REQUIRED: <base64: x402Version 2, amount 10000 (0.01 USDC), network eip155:8453> Where it is listed: - Coinbase's x402 Bazaar (the CDP discovery catalogue). All nine packs are listed. The Bazaar adds a resource only after a real payment has settled, so we bought each pack once ourselves. - x402scan, which registered all 37 of our paid resources from our OpenAPI file. - Our own manifest at /.well-known/x402. One thing we learned the hard way: directories probe bare routes without parameters. A route that answers 400 instead of 402 looks broken to them. Ours answers 402 with the published terms even when no id is given. 3. An agent-readable website - llms.txt: what we measure, when to call us, which call is free, and which pack costs what. - openapi.json: every endpoint, including the paid ones. - /.well-known/agent-card.json: an A2A agent card. - robots.txt explicitly welcomes retrieval and answer-engine crawlers. Until September it blocked our own API from them, which left AI answers quoting stale prose instead of live figures. 4. An identity: ERC-8004 Agentic Finance Graph is itself registered as ERC-8004 agent #95875 on Base. If we track agents' identities and wallets, we should have one too. 5. Open source on GitHub - agentic-finance-graph-mcp: the MCP server and CLI (Apache-2.0), with a security scan on every push. - agentic-finance-graph-spec: the open accounting spec behind every figure (CC BY 4.0). - Listed in hashgraph-online/awesome-ai-plugins. What we'd tell anyone listing a service for agents 1. Verify your name. Domain-verified registry names stop impostors and cost one TXT record or one well-known file. 2. Answer bare routes. Catalogues probe endpoints with no parameters. Give them a useful 402, never a confusing 400. 3. Settle once. Some x402 catalogues list a resource only after a real payment. Budget a few cents to buy your own products. 4. Keep a free path. Agents (and people) trust a paid pack more when the free record it packages is one call away. 5. Date and define every number. Agents quote you. Give them an as_of and a definition id to quote with it. Everything above is free to read under CC BY 4.0: agenticfinancegraph.com. The MCP server, the API and the spec are public. If you build agents that pay, tell us what you'd want to ask the ledger.
@TheInsumerModel mcp-server-insumer gives agents 27 condition-based access tools, signed attestations over 37 chains, via npx. Fits HOL Guard MCP contributions as mcp.insumer.json in contributions/mcp-servers/: add it, open a PR github.com/hashgraph-online/hol-guard/blob/main/C…
@dwhitedesign cuelume plays a sound when a long Claude Code turn ends so you can look away. Fits awesome-ai-plugins Tools & Integrations: add a one-sentence README entry and open your PR github.com/hashgraph-online/awesome-ai-plugins/bl…
@_overment Streamable-HTTP MCP server template updated for the latest spec. Fits awesome-ai-plugins Tools & Integrations: add a one-sentence README entry and open your PR github.com/hashgraph-online/awesome-ai-plugins/bl…
@olanetsoft Agent Lookout shows your running Claude Code and Codex sessions on one local page. Fits awesome-ai-plugins Tools & Integrations: add a one-sentence README entry and open your PR github.com/hashgraph-online/awesome-ai-plugins/bl…