feat: issue_close + issue_reopen tools (#30)

Adds two MCP tools that PATCH /api/v1/repos/{owner}/{name}/issues/{number}
with {"state":"closed"} or {"state":"open"}. Both use a shared
SetIssueState helper on the gitea client.

- internal/gitea/issues.go: SetIssueState method using the existing
  PatchJSON + MapStatus + json.Unmarshal pattern from GetIssue.
- internal/tools/issue_close.go: IssueClose tool. owner+name+number
  args. Owner allowlist enforced. Returns the updated issue. Reversible
  via issue_reopen, classified LOW risk.
- internal/tools/issue_reopen.go: mirror of IssueClose with
  state="open". Same risk profile.
- Registered both tools in cmd/gitea-mcp/main.go.
- Tests for both: success (asserts PATCH method, path, body), 404,
  and allowlist rejection — same shape as issue_get_test.go.

Closes #30

.aider.conventions.md

@@ -116,18 +116,64 @@ See `~/dev/PROJECT_SUMMARY.md` for detailed descriptions of each project.
 - **koala-ai-stack** (`AGENTS/`) — local AI server infrastructure management
 - **klimatkollen** (`XT/`) — Swedish municipal climate data platform
 
-## Knowledge base
+## Knowledge base — actively use it
 
-When available, agents can query the shared knowledge base:
+A persistent brain (BM25 search + LLM-synthesised Q&A) survives across sessions,
+hosts, and harnesses. It holds 100+ hard-won entries: infra incident postmortems,
+Go pitfalls, framework gotchas, design principles, ADRs. **It is not optional
+reference material — query it actively, not just when explicitly told.**
 
-- **MCP**: `mcp://hyperguild.<TAILNET>.ts.net:3100/knowledge`
-- **HTTP**: `http://hyperguild.<TAILNET>.ts.net:3100/api/v1/search`
+### When to query (treat as a reflex)
 
-<!-- TODO: replace <TAILNET> placeholder with the real Tailscale tailnet
-     name once hyperguild is deployed. Until then, agents that try to
-     reach the knowledge service on a host where it isn't running will
-     get DNS NXDOMAIN, which is the desired fail-loudly behavior. -->
-- **Scoping**: defaults to `public` collection; client projects filter to `{client}` + `public`
+- **Before** starting a non-trivial task — search for prior art with the symptom
+  AND the system component ("how did we solve X in Y?"). 5 seconds beats 5 hours.
+- **When debugging** — search for the error string, the stack frame, the affected
+  service. Past you may have already paid this tax.
+- **Before adopting** a pattern, library, framework, or model name — check if it
+  was tried and rejected, or what the integration footguns are.
+- **When making architectural decisions** — search for the domain + "ADR" or
+  "decision" to find prior reasoning before re-deriving it.
+- **When a recommendation feels novel** — challenge yourself: "has this been
+  documented?" The brain often has it.
+
+### When to write
+
+After you discover something that **future-you would forget** and that **isn't
+recoverable from the code, git log, or PR description alone**:
+
+- Bugs whose root cause is non-obvious and generalisable beyond this project.
+- Framework / library / model-name quirks that bit you and would bite anyone.
+- Design principles validated under fire (e.g. "every `_get` needs a `_list`").
+- Postmortems for incidents: what broke, why, how diagnosed, what to do next time.
+
+DON'T write project status, sprint progress, PR summaries, or "what I did this
+session" — those rot fast and the originals are in git/gitea anyway. Brain
+entries that age well are about *why*, *how to avoid*, and *what to do when*.
+
+### How to access (per harness)
+
+| Harness | Query | Write |
+|---------|-------|-------|
+| **Claude Code, Claude Desktop** | `brain_query` (BM25), `brain_answer` (LLM-synth + sources) MCP tools | `brain_write` MCP tool |
+| **Crush, Pi, Antigravity, other MCP-capable** | same MCP server: `ingestion-brain` (via the `mcp__*_brain__*` namespace once authenticated) | same |
+| **Anything HTTP-only (curl, scripts)** | `POST https://brain-mcp.d-ma.be/query` with `{"query":"..."}` (auth via `BRAIN_MCP_TOKEN`) | `POST .../write` with `{"content":"...","filename":"..."}` |
+| **Browser / human inspection** | `https://gitea.d-ma.be/mathias/hyperguild` → `knowledge/` and `wiki/` markdown files |
+
+- **Scoping**: defaults to `public` collection; client projects filter to `{client}` + `public`.
+- **Routing**: brain_answer's LLM uses berget.ai as primary, iguana ollama as
+  fallback. Both are configurable in the `supervisor/ingestion-deployment.yaml`
+  on the koala k3s cluster; don't hardcode local-only model names into the
+  berget URL (see knowledge entry on namespace mismatches).
+
+### Quick reflex checks
+
+If you find yourself about to say any of these out loud, you owe yourself a brain query first:
+
+- "I think the issue might be..."
+- "Let me try X and see..."
+- "I'll just write a script to..."
+- "This is probably a new bug..."
+- "Has anyone done this before?" — *yes, probably, go check.*
 
 ## Client work rules
 

.context/system-prompt.txt

@@ -121,18 +121,64 @@ See `~/dev/PROJECT_SUMMARY.md` for detailed descriptions of each project.
 - **koala-ai-stack** (`AGENTS/`) — local AI server infrastructure management
 - **klimatkollen** (`XT/`) — Swedish municipal climate data platform
 
-## Knowledge base
+## Knowledge base — actively use it
 
-When available, agents can query the shared knowledge base:
+A persistent brain (BM25 search + LLM-synthesised Q&A) survives across sessions,
+hosts, and harnesses. It holds 100+ hard-won entries: infra incident postmortems,
+Go pitfalls, framework gotchas, design principles, ADRs. **It is not optional
+reference material — query it actively, not just when explicitly told.**
 
-- **MCP**: `mcp://hyperguild.<TAILNET>.ts.net:3100/knowledge`
-- **HTTP**: `http://hyperguild.<TAILNET>.ts.net:3100/api/v1/search`
+### When to query (treat as a reflex)
 
-<!-- TODO: replace <TAILNET> placeholder with the real Tailscale tailnet
-     name once hyperguild is deployed. Until then, agents that try to
-     reach the knowledge service on a host where it isn't running will
-     get DNS NXDOMAIN, which is the desired fail-loudly behavior. -->
-- **Scoping**: defaults to `public` collection; client projects filter to `{client}` + `public`
+- **Before** starting a non-trivial task — search for prior art with the symptom
+  AND the system component ("how did we solve X in Y?"). 5 seconds beats 5 hours.
+- **When debugging** — search for the error string, the stack frame, the affected
+  service. Past you may have already paid this tax.
+- **Before adopting** a pattern, library, framework, or model name — check if it
+  was tried and rejected, or what the integration footguns are.
+- **When making architectural decisions** — search for the domain + "ADR" or
+  "decision" to find prior reasoning before re-deriving it.
+- **When a recommendation feels novel** — challenge yourself: "has this been
+  documented?" The brain often has it.
+
+### When to write
+
+After you discover something that **future-you would forget** and that **isn't
+recoverable from the code, git log, or PR description alone**:
+
+- Bugs whose root cause is non-obvious and generalisable beyond this project.
+- Framework / library / model-name quirks that bit you and would bite anyone.
+- Design principles validated under fire (e.g. "every `_get` needs a `_list`").
+- Postmortems for incidents: what broke, why, how diagnosed, what to do next time.
+
+DON'T write project status, sprint progress, PR summaries, or "what I did this
+session" — those rot fast and the originals are in git/gitea anyway. Brain
+entries that age well are about *why*, *how to avoid*, and *what to do when*.
+
+### How to access (per harness)
+
+| Harness | Query | Write |
+|---------|-------|-------|
+| **Claude Code, Claude Desktop** | `brain_query` (BM25), `brain_answer` (LLM-synth + sources) MCP tools | `brain_write` MCP tool |
+| **Crush, Pi, Antigravity, other MCP-capable** | same MCP server: `ingestion-brain` (via the `mcp__*_brain__*` namespace once authenticated) | same |
+| **Anything HTTP-only (curl, scripts)** | `POST https://brain-mcp.d-ma.be/query` with `{"query":"..."}` (auth via `BRAIN_MCP_TOKEN`) | `POST .../write` with `{"content":"...","filename":"..."}` |
+| **Browser / human inspection** | `https://gitea.d-ma.be/mathias/hyperguild` → `knowledge/` and `wiki/` markdown files |
+
+- **Scoping**: defaults to `public` collection; client projects filter to `{client}` + `public`.
+- **Routing**: brain_answer's LLM uses berget.ai as primary, iguana ollama as
+  fallback. Both are configurable in the `supervisor/ingestion-deployment.yaml`
+  on the koala k3s cluster; don't hardcode local-only model names into the
+  berget URL (see knowledge entry on namespace mismatches).
+
+### Quick reflex checks
+
+If you find yourself about to say any of these out loud, you owe yourself a brain query first:
+
+- "I think the issue might be..."
+- "Let me try X and see..."
+- "I'll just write a script to..."
+- "This is probably a new bug..."
+- "Has anyone done this before?" — *yes, probably, go check.*
 
 ## Client work rules
 

.cursorrules

@@ -119,18 +119,64 @@ See `~/dev/PROJECT_SUMMARY.md` for detailed descriptions of each project.
 - **koala-ai-stack** (`AGENTS/`) — local AI server infrastructure management
 - **klimatkollen** (`XT/`) — Swedish municipal climate data platform
 
-## Knowledge base
+## Knowledge base — actively use it
 
-When available, agents can query the shared knowledge base:
+A persistent brain (BM25 search + LLM-synthesised Q&A) survives across sessions,
+hosts, and harnesses. It holds 100+ hard-won entries: infra incident postmortems,
+Go pitfalls, framework gotchas, design principles, ADRs. **It is not optional
+reference material — query it actively, not just when explicitly told.**
 
-- **MCP**: `mcp://hyperguild.<TAILNET>.ts.net:3100/knowledge`
-- **HTTP**: `http://hyperguild.<TAILNET>.ts.net:3100/api/v1/search`
+### When to query (treat as a reflex)
 
-<!-- TODO: replace <TAILNET> placeholder with the real Tailscale tailnet
-     name once hyperguild is deployed. Until then, agents that try to
-     reach the knowledge service on a host where it isn't running will
-     get DNS NXDOMAIN, which is the desired fail-loudly behavior. -->
-- **Scoping**: defaults to `public` collection; client projects filter to `{client}` + `public`
+- **Before** starting a non-trivial task — search for prior art with the symptom
+  AND the system component ("how did we solve X in Y?"). 5 seconds beats 5 hours.
+- **When debugging** — search for the error string, the stack frame, the affected
+  service. Past you may have already paid this tax.
+- **Before adopting** a pattern, library, framework, or model name — check if it
+  was tried and rejected, or what the integration footguns are.
+- **When making architectural decisions** — search for the domain + "ADR" or
+  "decision" to find prior reasoning before re-deriving it.
+- **When a recommendation feels novel** — challenge yourself: "has this been
+  documented?" The brain often has it.
+
+### When to write
+
+After you discover something that **future-you would forget** and that **isn't
+recoverable from the code, git log, or PR description alone**:
+
+- Bugs whose root cause is non-obvious and generalisable beyond this project.
+- Framework / library / model-name quirks that bit you and would bite anyone.
+- Design principles validated under fire (e.g. "every `_get` needs a `_list`").
+- Postmortems for incidents: what broke, why, how diagnosed, what to do next time.
+
+DON'T write project status, sprint progress, PR summaries, or "what I did this
+session" — those rot fast and the originals are in git/gitea anyway. Brain
+entries that age well are about *why*, *how to avoid*, and *what to do when*.
+
+### How to access (per harness)
+
+| Harness | Query | Write |
+|---------|-------|-------|
+| **Claude Code, Claude Desktop** | `brain_query` (BM25), `brain_answer` (LLM-synth + sources) MCP tools | `brain_write` MCP tool |
+| **Crush, Pi, Antigravity, other MCP-capable** | same MCP server: `ingestion-brain` (via the `mcp__*_brain__*` namespace once authenticated) | same |
+| **Anything HTTP-only (curl, scripts)** | `POST https://brain-mcp.d-ma.be/query` with `{"query":"..."}` (auth via `BRAIN_MCP_TOKEN`) | `POST .../write` with `{"content":"...","filename":"..."}` |
+| **Browser / human inspection** | `https://gitea.d-ma.be/mathias/hyperguild` → `knowledge/` and `wiki/` markdown files |
+
+- **Scoping**: defaults to `public` collection; client projects filter to `{client}` + `public`.
+- **Routing**: brain_answer's LLM uses berget.ai as primary, iguana ollama as
+  fallback. Both are configurable in the `supervisor/ingestion-deployment.yaml`
+  on the koala k3s cluster; don't hardcode local-only model names into the
+  berget URL (see knowledge entry on namespace mismatches).
+
+### Quick reflex checks
+
+If you find yourself about to say any of these out loud, you owe yourself a brain query first:
+
+- "I think the issue might be..."
+- "Let me try X and see..."
+- "I'll just write a script to..."
+- "This is probably a new bug..."
+- "Has anyone done this before?" — *yes, probably, go check.*
 
 ## Client work rules
 

AGENTS.md

@@ -116,18 +116,64 @@ See `~/dev/PROJECT_SUMMARY.md` for detailed descriptions of each project.
 - **koala-ai-stack** (`AGENTS/`) — local AI server infrastructure management
 - **klimatkollen** (`XT/`) — Swedish municipal climate data platform
 
-## Knowledge base
+## Knowledge base — actively use it
 
-When available, agents can query the shared knowledge base:
+A persistent brain (BM25 search + LLM-synthesised Q&A) survives across sessions,
+hosts, and harnesses. It holds 100+ hard-won entries: infra incident postmortems,
+Go pitfalls, framework gotchas, design principles, ADRs. **It is not optional
+reference material — query it actively, not just when explicitly told.**
 
-- **MCP**: `mcp://hyperguild.<TAILNET>.ts.net:3100/knowledge`
-- **HTTP**: `http://hyperguild.<TAILNET>.ts.net:3100/api/v1/search`
+### When to query (treat as a reflex)
 
-<!-- TODO: replace <TAILNET> placeholder with the real Tailscale tailnet
-     name once hyperguild is deployed. Until then, agents that try to
-     reach the knowledge service on a host where it isn't running will
-     get DNS NXDOMAIN, which is the desired fail-loudly behavior. -->
-- **Scoping**: defaults to `public` collection; client projects filter to `{client}` + `public`
+- **Before** starting a non-trivial task — search for prior art with the symptom
+  AND the system component ("how did we solve X in Y?"). 5 seconds beats 5 hours.
+- **When debugging** — search for the error string, the stack frame, the affected
+  service. Past you may have already paid this tax.
+- **Before adopting** a pattern, library, framework, or model name — check if it
+  was tried and rejected, or what the integration footguns are.
+- **When making architectural decisions** — search for the domain + "ADR" or
+  "decision" to find prior reasoning before re-deriving it.
+- **When a recommendation feels novel** — challenge yourself: "has this been
+  documented?" The brain often has it.
+
+### When to write
+
+After you discover something that **future-you would forget** and that **isn't
+recoverable from the code, git log, or PR description alone**:
+
+- Bugs whose root cause is non-obvious and generalisable beyond this project.
+- Framework / library / model-name quirks that bit you and would bite anyone.
+- Design principles validated under fire (e.g. "every `_get` needs a `_list`").
+- Postmortems for incidents: what broke, why, how diagnosed, what to do next time.
+
+DON'T write project status, sprint progress, PR summaries, or "what I did this
+session" — those rot fast and the originals are in git/gitea anyway. Brain
+entries that age well are about *why*, *how to avoid*, and *what to do when*.
+
+### How to access (per harness)
+
+| Harness | Query | Write |
+|---------|-------|-------|
+| **Claude Code, Claude Desktop** | `brain_query` (BM25), `brain_answer` (LLM-synth + sources) MCP tools | `brain_write` MCP tool |
+| **Crush, Pi, Antigravity, other MCP-capable** | same MCP server: `ingestion-brain` (via the `mcp__*_brain__*` namespace once authenticated) | same |
+| **Anything HTTP-only (curl, scripts)** | `POST https://brain-mcp.d-ma.be/query` with `{"query":"..."}` (auth via `BRAIN_MCP_TOKEN`) | `POST .../write` with `{"content":"...","filename":"..."}` |
+| **Browser / human inspection** | `https://gitea.d-ma.be/mathias/hyperguild` → `knowledge/` and `wiki/` markdown files |
+
+- **Scoping**: defaults to `public` collection; client projects filter to `{client}` + `public`.
+- **Routing**: brain_answer's LLM uses berget.ai as primary, iguana ollama as
+  fallback. Both are configurable in the `supervisor/ingestion-deployment.yaml`
+  on the koala k3s cluster; don't hardcode local-only model names into the
+  berget URL (see knowledge entry on namespace mismatches).
+
+### Quick reflex checks
+
+If you find yourself about to say any of these out loud, you owe yourself a brain query first:
+
+- "I think the issue might be..."
+- "Let me try X and see..."
+- "I'll just write a script to..."
+- "This is probably a new bug..."
+- "Has anyone done this before?" — *yes, probably, go check.*
 
 ## Client work rules
 

cmd/gitea-mcp/main.go

@@ -66,6 +66,8 @@ func main() {
 	reg.Register(tools.NewRepoTree(giteaClient, ownerAllow))
 	reg.Register(tools.NewRepoTopicsUpdate(giteaClient, ownerAllow))
 	reg.Register(tools.NewIssueGet(giteaClient, ownerAllow))
+	reg.Register(tools.NewIssueClose(giteaClient, ownerAllow))
+	reg.Register(tools.NewIssueReopen(giteaClient, ownerAllow))
 	reg.Register(tools.NewReleaseCreate(giteaClient, ownerAllow))
 	reg.Register(tools.NewRepoDelete(giteaClient, ownerAllow))
 

internal/gitea/issues.go

@@ -72,6 +72,28 @@ func (c *Client) CreateIssue(ctx context.Context, owner, repo string, args Creat
 	return &iss, nil
 }
 
+// SetIssueState flips an issue between "open" and "closed" via PATCH.
+// Gitea uses the same endpoint for both transitions.
+func (c *Client) SetIssueState(ctx context.Context, owner, repo string, number int, state string) (*Issue, error) {
+	p := fmt.Sprintf("/api/v1/repos/%s/%s/issues/%d", owner, repo, number)
+	payload, err := json.Marshal(map[string]string{"state": state})
+	if err != nil {
+		return nil, err
+	}
+	body, status, err := c.PatchJSON(ctx, p, payload)
+	if err != nil {
+		return nil, err
+	}
+	if err := MapStatus(status, body); err != nil {
+		return nil, err
+	}
+	var iss Issue
+	if err := json.Unmarshal(body, &iss); err != nil {
+		return nil, err
+	}
+	return &iss, nil
+}
+
 type IssueComment struct {
 	ID      int64  `json:"id"`
 	Body    string `json:"body"`

internal/tools/issue_close.go

@@ -0,0 +1,56 @@
+package tools
+
+import (
+	"context"
+	"encoding/json"
+
+	"gitea.d-ma.be/mathias/gitea-mcp/internal/allowlist"
+	"gitea.d-ma.be/mathias/gitea-mcp/internal/gitea"
+	"gitea.d-ma.be/mathias/gitea-mcp/internal/registry"
+)
+
+type IssueClose struct {
+	c *gitea.Client
+	a *allowlist.Allowlist
+}
+
+func NewIssueClose(c *gitea.Client, a *allowlist.Allowlist) *IssueClose {
+	return &IssueClose{c: c, a: a}
+}
+
+func (t *IssueClose) Descriptor() registry.ToolDescriptor {
+	return registry.ToolDescriptor{
+		Name:        "issue_close",
+		Description: "Close an open issue. Reversible via issue_reopen.",
+		InputSchema: json.RawMessage(`{
+			"type":"object",
+			"properties":{
+				"owner":{"type":"string"},
+				"name":{"type":"string"},
+				"number":{"type":"integer","minimum":1}
+			},
+			"required":["owner","name","number"]
+		}`),
+	}
+}
+
+type issueCloseArgs struct {
+	Owner  string `json:"owner"`
+	Name   string `json:"name"`
+	Number int    `json:"number"`
+}
+
+func (t *IssueClose) Call(ctx context.Context, raw json.RawMessage) (json.RawMessage, error) {
+	var args issueCloseArgs
+	if err := parseArgs(raw, &args); err != nil {
+		return nil, err
+	}
+	if err := t.a.Check(args.Owner); err != nil {
+		return nil, err
+	}
+	iss, err := t.c.SetIssueState(ctx, args.Owner, args.Name, args.Number, "closed")
+	if err != nil {
+		return nil, err
+	}
+	return textOK(iss)
+}

internal/tools/issue_close_test.go

@@ -0,0 +1,52 @@
+package tools_test
+
+import (
+	"context"
+	"encoding/json"
+	"io"
+	"net/http"
+	"net/http/httptest"
+	"testing"
+
+	"gitea.d-ma.be/mathias/gitea-mcp/internal/allowlist"
+	"gitea.d-ma.be/mathias/gitea-mcp/internal/gitea"
+	"gitea.d-ma.be/mathias/gitea-mcp/internal/tools"
+	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
+)
+
+func TestIssueCloseTool(t *testing.T) {
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		assert.Equal(t, http.MethodPatch, r.Method)
+		assert.Equal(t, "/api/v1/repos/mathias/infra/issues/26", r.URL.Path)
+		b, _ := io.ReadAll(r.Body)
+		assert.JSONEq(t, `{"state":"closed"}`, string(b))
+		w.Header().Set("Content-Type", "application/json")
+		_, _ = w.Write([]byte(`{"number":26,"title":"feat: ntfy via NPM","state":"closed","html_url":"http://gitea.example.com/mathias/infra/issues/26"}`))
+	}))
+	defer srv.Close()
+
+	tool := tools.NewIssueClose(gitea.NewClient(srv.URL, "tok"), allowlist.New([]string{"mathias"}))
+	out, err := tool.Call(context.Background(), json.RawMessage(`{"owner":"mathias","name":"infra","number":26}`))
+	require.NoError(t, err)
+	assert.Contains(t, string(out), `"number":26`)
+	assert.Contains(t, string(out), `"state":"closed"`)
+}
+
+func TestIssueCloseTool_NotFound(t *testing.T) {
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		w.WriteHeader(http.StatusNotFound)
+		_, _ = w.Write([]byte(`{"message":"issue not found"}`))
+	}))
+	defer srv.Close()
+
+	tool := tools.NewIssueClose(gitea.NewClient(srv.URL, "tok"), allowlist.New([]string{"mathias"}))
+	_, err := tool.Call(context.Background(), json.RawMessage(`{"owner":"mathias","name":"infra","number":999}`))
+	require.Error(t, err)
+}
+
+func TestIssueCloseAllowlistRejects(t *testing.T) {
+	tool := tools.NewIssueClose(gitea.NewClient("http://unused", ""), allowlist.New([]string{"mathias"}))
+	_, err := tool.Call(context.Background(), json.RawMessage(`{"owner":"evil","name":"x","number":1}`))
+	require.Error(t, err)
+}

internal/tools/issue_reopen.go

@@ -0,0 +1,56 @@
+package tools
+
+import (
+	"context"
+	"encoding/json"
+
+	"gitea.d-ma.be/mathias/gitea-mcp/internal/allowlist"
+	"gitea.d-ma.be/mathias/gitea-mcp/internal/gitea"
+	"gitea.d-ma.be/mathias/gitea-mcp/internal/registry"
+)
+
+type IssueReopen struct {
+	c *gitea.Client
+	a *allowlist.Allowlist
+}
+
+func NewIssueReopen(c *gitea.Client, a *allowlist.Allowlist) *IssueReopen {
+	return &IssueReopen{c: c, a: a}
+}
+
+func (t *IssueReopen) Descriptor() registry.ToolDescriptor {
+	return registry.ToolDescriptor{
+		Name:        "issue_reopen",
+		Description: "Reopen a closed issue. Reversible via issue_close.",
+		InputSchema: json.RawMessage(`{
+			"type":"object",
+			"properties":{
+				"owner":{"type":"string"},
+				"name":{"type":"string"},
+				"number":{"type":"integer","minimum":1}
+			},
+			"required":["owner","name","number"]
+		}`),
+	}
+}
+
+type issueReopenArgs struct {
+	Owner  string `json:"owner"`
+	Name   string `json:"name"`
+	Number int    `json:"number"`
+}
+
+func (t *IssueReopen) Call(ctx context.Context, raw json.RawMessage) (json.RawMessage, error) {
+	var args issueReopenArgs
+	if err := parseArgs(raw, &args); err != nil {
+		return nil, err
+	}
+	if err := t.a.Check(args.Owner); err != nil {
+		return nil, err
+	}
+	iss, err := t.c.SetIssueState(ctx, args.Owner, args.Name, args.Number, "open")
+	if err != nil {
+		return nil, err
+	}
+	return textOK(iss)
+}

internal/tools/issue_reopen_test.go

@@ -0,0 +1,40 @@
+package tools_test
+
+import (
+	"context"
+	"encoding/json"
+	"io"
+	"net/http"
+	"net/http/httptest"
+	"testing"
+
+	"gitea.d-ma.be/mathias/gitea-mcp/internal/allowlist"
+	"gitea.d-ma.be/mathias/gitea-mcp/internal/gitea"
+	"gitea.d-ma.be/mathias/gitea-mcp/internal/tools"
+	"github.com/stretchr/testify/assert"
+	"github.com/stretchr/testify/require"
+)
+
+func TestIssueReopenTool(t *testing.T) {
+	srv := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
+		assert.Equal(t, http.MethodPatch, r.Method)
+		assert.Equal(t, "/api/v1/repos/mathias/infra/issues/26", r.URL.Path)
+		b, _ := io.ReadAll(r.Body)
+		assert.JSONEq(t, `{"state":"open"}`, string(b))
+		w.Header().Set("Content-Type", "application/json")
+		_, _ = w.Write([]byte(`{"number":26,"title":"feat: ntfy via NPM","state":"open","html_url":"http://gitea.example.com/mathias/infra/issues/26"}`))
+	}))
+	defer srv.Close()
+
+	tool := tools.NewIssueReopen(gitea.NewClient(srv.URL, "tok"), allowlist.New([]string{"mathias"}))
+	out, err := tool.Call(context.Background(), json.RawMessage(`{"owner":"mathias","name":"infra","number":26}`))
+	require.NoError(t, err)
+	assert.Contains(t, string(out), `"number":26`)
+	assert.Contains(t, string(out), `"state":"open"`)
+}
+
+func TestIssueReopenAllowlistRejects(t *testing.T) {
+	tool := tools.NewIssueReopen(gitea.NewClient("http://unused", ""), allowlist.New([]string{"mathias"}))
+	_, err := tool.Call(context.Background(), json.RawMessage(`{"owner":"evil","name":"x","number":1}`))
+	require.Error(t, err)
+}