[PR #126] [CLOSED] fix: handle large deliverables with chunked writing and audit-log copying #130

New issue

Closed

opened 2026-02-27 08:09:23 +03:00 by kerem · 0 comments

kerem commented

2026-02-27 08:09:23 +03:00

Owner

📋 Pull Request Information

Original PR: https://github.com/KeygraphHQ/shannon/pull/126
Author: @ajmallesh
Created: 2/12/2026
Status: ❌ Closed

Base: main ← Head: fix/large-deliverable-handling

📝 Commits (8)

f7eb3a5 fix: add file_path parameter to save_deliverable for large reports
4ea255b docs: simplify and condense CLAUDE.md
efc492a feat: add issue number detection to pr command
57d82b4 chore: remove CLAUDE_CODE_MAX_OUTPUT_TOKENS env var handling
a3ba152 fix: restore max_output_tokens error handling
c90993e fix: add chunked writing instructions to all agent prompts
ef7b4dd feat: copy deliverables to audit-logs for self-contained audit trail
b7c48ab fix: restore CLAUDE_CODE_MAX_OUTPUT_TOKENS env var support

📊 Changes

22 files changed (+408 additions, -335 deletions)

View changed files

📝 .claude/commands/pr.md (+18 -0)
📝 .env.example (+3 -0)
📝 CLAUDE.md (+92 -280)
📝 README.md (+2 -0)
📝 docker-compose.yml (+1 -0)
📝 mcp-server/src/tools/save-deliverable.ts (+73 -19)
📝 prompts/exploit-auth.txt (+11 -2)
📝 prompts/exploit-authz.txt (+11 -2)
📝 prompts/exploit-injection.txt (+11 -2)
📝 prompts/exploit-ssrf.txt (+11 -2)
📝 prompts/exploit-xss.txt (+12 -4)
📝 prompts/pre-recon-code.txt (+13 -4)
📝 prompts/recon.txt (+15 -2)
📝 prompts/vuln-auth.txt (+12 -3)
📝 prompts/vuln-authz.txt (+12 -3)
📝 prompts/vuln-injection.txt (+12 -3)
📝 prompts/vuln-ssrf.txt (+12 -3)
📝 prompts/vuln-xss.txt (+12 -3)
📝 src/ai/claude-executor.ts (+13 -0)
📝 src/ai/message-handlers.ts (+18 -1)

...and 2 more files

📄 Description

Summary

Add chunked writing instructions to all agent prompts for handling large reports that exceed output limits
Copy deliverables to audit-logs directory for self-contained audit trails
Restore CLAUDE_CODE_MAX_OUTPUT_TOKENS env var support with proper error handling
Add file_path parameter to save_deliverable MCP tool for large reports

_{🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.}

## 📋 Pull Request Information **Original PR:** https://github.com/KeygraphHQ/shannon/pull/126 **Author:** [@ajmallesh](https://github.com/ajmallesh) **Created:** 2/12/2026 **Status:** ❌ Closed **Base:** `main` ← **Head:** `fix/large-deliverable-handling` --- ### 📝 Commits (8) - [`f7eb3a5`](https://github.com/KeygraphHQ/shannon/commit/f7eb3a57cec88f0c96a7499dced29ad6c0ce7823) fix: add file_path parameter to save_deliverable for large reports - [`4ea255b`](https://github.com/KeygraphHQ/shannon/commit/4ea255b8c542482647c7978ac4e52f228cda290f) docs: simplify and condense CLAUDE.md - [`efc492a`](https://github.com/KeygraphHQ/shannon/commit/efc492a3f9795ca2fb9a13390fff605b0d0f4129) feat: add issue number detection to pr command - [`57d82b4`](https://github.com/KeygraphHQ/shannon/commit/57d82b41aebbc0ef9d2fe5a6d999ab2d7102bb08) chore: remove CLAUDE_CODE_MAX_OUTPUT_TOKENS env var handling - [`a3ba152`](https://github.com/KeygraphHQ/shannon/commit/a3ba1529e0989e15215f30ce8625eb1d9b1471fa) fix: restore max_output_tokens error handling - [`c90993e`](https://github.com/KeygraphHQ/shannon/commit/c90993e1bc4fa6c8aa2d3c07b2fa657c7c8b10b0) fix: add chunked writing instructions to all agent prompts - [`ef7b4dd`](https://github.com/KeygraphHQ/shannon/commit/ef7b4ddf301e35de78008363e72c2605ba891f87) feat: copy deliverables to audit-logs for self-contained audit trail - [`b7c48ab`](https://github.com/KeygraphHQ/shannon/commit/b7c48abd0229edf5c2c663b74a89b3a136b1bf92) fix: restore CLAUDE_CODE_MAX_OUTPUT_TOKENS env var support ### 📊 Changes **22 files changed** (+408 additions, -335 deletions) <details> <summary>View changed files</summary> 📝 `.claude/commands/pr.md` (+18 -0) 📝 `.env.example` (+3 -0) 📝 `CLAUDE.md` (+92 -280) 📝 `README.md` (+2 -0) 📝 `docker-compose.yml` (+1 -0) 📝 `mcp-server/src/tools/save-deliverable.ts` (+73 -19) 📝 `prompts/exploit-auth.txt` (+11 -2) 📝 `prompts/exploit-authz.txt` (+11 -2) 📝 `prompts/exploit-injection.txt` (+11 -2) 📝 `prompts/exploit-ssrf.txt` (+11 -2) 📝 `prompts/exploit-xss.txt` (+12 -4) 📝 `prompts/pre-recon-code.txt` (+13 -4) 📝 `prompts/recon.txt` (+15 -2) 📝 `prompts/vuln-auth.txt` (+12 -3) 📝 `prompts/vuln-authz.txt` (+12 -3) 📝 `prompts/vuln-injection.txt` (+12 -3) 📝 `prompts/vuln-ssrf.txt` (+12 -3) 📝 `prompts/vuln-xss.txt` (+12 -3) 📝 `src/ai/claude-executor.ts` (+13 -0) 📝 `src/ai/message-handlers.ts` (+18 -1) _...and 2 more files_ </details> ### 📄 Description ## Summary - Add chunked writing instructions to all agent prompts for handling large reports that exceed output limits - Copy deliverables to audit-logs directory for self-contained audit trails - Restore `CLAUDE_CODE_MAX_OUTPUT_TOKENS` env var support with proper error handling - Add `file_path` parameter to `save_deliverable` MCP tool for large reports --- <sub>🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.</sub>

kerem

2026-02-27 08:09:23 +03:00

closed this issue
added the
pull-request
label

No labels

pull-request

No milestone

No project

No assignees

1 participant

Notifications

Due date

The due date is invalid or out of range. Please use the format "yyyy-mm-dd".

No due date set.

Dependencies

No dependencies set.

Reference

starred/shannon-KeygraphHQ#130

No description provided.

Rows
Columns